New perspectives in international business research 9781848552791, 1848552793

Table of contents :
New perspectives in IB research: culture, governance, Entrepreneurship and international expansion. Headquarters-subsidiary relationships and the country-of-origin effect. An exploratory study of culture distance and perceptions of relational risk. Drivers of interpersonal and inter-unit trust in multinational corporations. Normative control in MNC. Exploring the antecedents of relationship commitment in an import-export dyad. Constructing jurisdictional advantage. Comparative international entrepreneurship: The software industry in the Indian sub-continent. Reinterpreting a `prime example' of a born global: Cochlear's international launch. Myths in microfinance. R&D and foreign direct investment with asymmetries in knowledge transmission. FDI and spillovers in the Swiss manufacturing industry: Interaction effects between spillover mechanisms and domestic absorptive capacities. Market and technology access through firm acquisitions: Beyond one size fits all. List of Contributors. Progress in international business research. New perspectives in international business research. Copyright page.

Citation preview

NEW PERSPECTIVES IN INTERNATIONAL BUSINESS RESEARCH

PROGRESS IN INTERNATIONAL BUSINESS RESEARCH Series Editor: The European International Business Academy (EIBA) Recent Volumes: Volume 1:

Progress in International Business Research – Edited by Gabriel R. G. Benito and Henrich R. Greve

Volume 2:

Foreign Direct Investment, Location and Competitiveness – Edited by John H. Dunning and Philippe Gugler

PROGRESS IN INTERNATIONAL BUSINESS RESEARCH VOLUME 3

NEW PERSPECTIVES IN INTERNATIONAL BUSINESS RESEARCH EDITED BY

MARYANN P. FELDMAN University of North Carolina, NC, USA

GRAZIA D. SANTANGELO Universita` degli Studi di Catania, Catania, Italy

United Kingdom – North America – Japan India – Malaysia – China

JAI Press is an imprint of Emerald Group Publishing Limited Howard House, Wagon Lane, Bingley BD16 1WA, UK First edition 2008 Copyright r 2008 Emerald Group Publishing Limited Reprints and permission service Contact: [email protected] No part of this book may be reproduced, stored in a retrieval system, transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without either the prior written permission of the publisher or a licence permitting restricted copying issued in the UK by The Copyright Licensing Agency and in the USA by The Copyright Clearance Center. No responsibility is accepted for the accuracy of information contained in the text, illustrations or advertisements. The opinions expressed in these chapters are not necessarily those of the Editor or the publisher. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN: 978-1-84855-278-4 ISSN: 1745-8862 (Series)

Awarded in recognition of Emerald’s production department’s adherence to quality systems and processes when preparing scholarly journals for print

CONTENTS LIST OF CONTRIBUTORS

ix

NEW PERSPECTIVES IN IB RESEARCH: CULTURE, GOVERNANCE, ENTREPRENEURSHIP AND INTERNATIONAL EXPANSION Maryann Feldman and Grazia D. Santangelo

1

PART I: MANAGING CULTURAL DIFFERENCES ACROSS COUNTRIES HEADQUARTERS–SUBSIDIARY RELATIONSHIPS AND THE COUNTRY-OF-ORIGIN EFFECT Anne-Wil Harzing and Niels Noorderhaven

13

AN EXPLORATORY STUDY OF CULTURE DISTANCE AND PERCEPTIONS OF RELATIONAL RISK Susana Costa e Silva and Luciara Nardon

41

DRIVERS OF INTERPERSONAL AND INTER-UNIT TRUST IN MULTINATIONAL CORPORATIONS Kristiina Ma¨kela¨, Wilhelm Barner-Rasmussen and Ingmar Bjo¨rkman

59

PART II: GOVERNANCE NORMATIVE CONTROL IN MNCS: A MICRO-POLITICAL CONFLICT PERSPECTIVE Susanne Blazejewski v

83

vi

CONTENTS

EXPLORING THE ANTECEDENTS OF RELATIONSHIP COMMITMENT IN AN IMPORT–EXPORT DYAD Diana Weinberg and Abraham Carmeli

113

CONSTRUCTING JURISDICTIONAL ADVANTAGE Maryann Feldman and Roger Martin

139

PART III: THE GENESIS OF INTERNATIONAL ENTREPRENEURS COMPARATIVE INTERNATIONAL ENTREPRENEURSHIP: THE SOFTWARE INDUSTRY IN THE INDIAN SUB-CONTINENT Shameen Prashantham, Amer Qureshi and Stephen Young

165

REINTERPRETING A ‘PRIME EXAMPLE’ OF A BORN GLOBAL: COCHLEAR’S INTERNATIONAL LAUNCH Lisa Hewerdine and Catherine Welch

189

MYTHS IN MICROFINANCE Roy Mersland and R. Øystein Strøm

207

PART IV: TECHNOLOGY AND INTERNATIONAL EXPANSION R&D AND FOREIGN DIRECT INVESTMENT WITH ASYMMETRIES IN KNOWLEDGE TRANSMISSION Maria Luisa Petit, Francesca Sanna-Randaccio and Roberta Sestini

231

Contents

vii

FDI AND SPILLOVERS IN THE SWISS MANUFACTURING INDUSTRY: INTERACTION EFFECTS BETWEEN SPILLOVER MECHANISMS AND DOMESTIC ABSORPTIVE CAPACITIES Lamia Ben Hamida and Philippe Gugler

263

MARKET AND TECHNOLOGY ACCESS THROUGH FIRM ACQUISITIONS: BEYOND ONE SIZE FITS ALL Christoph Grimpe and Katrin Hussinger

289

This page intentionally left blank

LIST OF CONTRIBUTORS Wilhelm Barner-Rasmussen

Department of Management and Organization, Hanken School of Economics, Helsinki, Finland

Lamia Ben Hamida

University of Fribourg and World Trade Institute, Berne

Ingmar Bjo¨rkman

Department of Management and Organization, Hanken School of Economics, Helsinki, Finland

Susanne Blazejewski

Department of Management, European University Viadrina, Frankfurt (Oder), Germany

Abraham Carmeli

Graduate School of Business Administration, Bar–Ilan University, Israel

Susana Costa e Silva

Faculdade de Economia e Gesta˜o Universidade Cato´lica Portuguesa Porto, Portugal

Maryann Feldman

Department of Public Policy, University of North Carolina, Chapel Hill, North Carolina

Christoph Grimpe

Centre for European Economic Research (ZEW), Mannheim, Germany

Philippe Gugler

Faculty of Economics and Social Sciences, University of Fribourg and World Trade Institute, Berne

ix

x

LIST OF CONTRIBUTORS

Anne-Wil Harzing

Department of Management and Marketing, University of Melbourne, Melbourne, Australia

Lisa Hewerdine

Business School, University of Adelaide, Australia

Katrin Hussinger

Department of Organization and Strategy, University of Maastricht, Maastricht, The Netherlands

Maria Luisa Petit

Department of Computer and Systems Sciences, Universita` di Roma La Sapienza, Rome, Italy

Kristiina Ma¨kela¨

Department of Management and Organization, Hanken School of Economics, Helsinki, Finland

Roger Martin

Rotman School of Management, University of Toronto, Ontario, Canada

Roy Mersland

University of Agder, Norway

Luciara Nardon

Charles H. Lundquist College of Business, University of Oregon, Eugene, USA

Niels Noorderhaven

Department of Organization & Strategy, Tilburg University, Tiburg, The Netherlands

Shameen Prashantham

CIER, Department of Management, University of Glasgow, Scotland, United Kingdom

Amer Qureshi

Q Consulting and Training, Lahore, Pakistan

Francesca Sanna-Randaccio

Department of Computer and Systems Sciences, University of Rome La Sapienza, Rome, Italy

Grazia D. Santangelo

Facolta` di Scienze Politiche, Universita` degli Studi di Catania, Catania, Italy

xi

List of Contributors

Roberta Sestini

Department of Computer and Systems Sciences, Universita` di Roma La Sapienza, Rome, Italy

Øystein Strøm

Østfold University College, Halden, Norway

Diana Weinberg

Graduate School of Business Administration, Bar–Ilan University, Israel

Catherine Welch

University of Sydney, Sydney, Australia

Stephen Young

CIER, Department of Management, University of Glasgow, Scotland, United Kingdom

This page intentionally left blank

NEW PERSPECTIVES IN IB RESEARCH: CULTURE, GOVERNANCE, ENTREPRENEURSHIP AND INTERNATIONAL EXPANSION Maryann Feldman and Grazia D. Santangelo 1. AIM OF THE VOLUME This volume is the outcome of the 33rd European International Business Academy (EIBA) conference held at the Faculty of Political Science of the University of Catania (Italy). This conference brought together more than 300 scholars from around the world to discuss theoretical and empirical issues in international business (IB), as well as their consequences and challenges to IB scholars and policy-makers. Organized around 10 thematic tracks, the conference is the annual forum for discussing major research issues in the IB realm. This volume is a collection of the best papers, which, selected through a blind refereeing process for presentation at the conference, make significant contributions by providing fresh new perspectives on a variety of relevant topics. After 33 years, the IB field is maturing to consider new issues related to culture, governance and entrepreneurship. Besides the traditional topics

New Perspectives in International Business Research Progress in International Business Research, Volume 3, 1–9 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1745-8862/doi:10.1016/S1745-8862(08)03001-X

1

2

MARYANN FELDMAN AND GRAZIA D. SANTANGELO

of multinationals’ investment decisions and FDI spillovers, this volume mirrors the recent evolution in the field by considering cultural differences in headquarter–subsidiaries relationships, perception differences of relational risk, and determinants of trust in interpersonal and inter-unit relationships. In addition, the governance of relationships within firms as in conflicts between parent–subsidiaries, between firms as in import–export dyads, and within locations as concerned the construction of jurisdictional advantage. We further consider the genesis of international entrepreneurs in the context of the software industry in India, the related debate on the definition of ‘born global’ firms and the controversies about the efficacy of microfinance to support third world development. Although these are just a few of the new perspectives rising within the IB realm, these chapters were selected to provide a coherent collection that demonstrates the vitality of research conducting in the field, IB’s responsiveness to new emerging topics in related areas and the fruitful incorporation of new ideas.

2. CHANGES IN THE INTERNATIONAL COMPETITIVE ENVIRONMENT AND NEW PERSPECTIVES IN IB RESEARCH The broadening-out of IB research to fresh new perspectives has enlarged the field of research to a wider range of themes that are relevant to the changing international competitive environment. Not only does the international competitive environment embody competition amongst firms, but there is also competition between cultures and institutional systems (Leung, Bhagat, Buchan, Erez, & Gibson, 2005 ; Ferner, 1997; Harzing & Sorge, 2003; and Noorderhaven & Harzing, 2003). Some firms have proved successful in achieving growth by promoting cooperation and openness; others have lost out and are beginning to emulate more successful firms. Yet the question of why some firms succeed while others fail is often due to differences in culture, which drives trust in relationships (Bolino, Turnley, & Bloodgood, 2002; Kostova & Roth, 2002) and also perceptions about acceptable risk and appropriate management strategy (Kluckhohn, 1954; Hofstede, 1980, 1991). A significant research question for IB scholars is then about how are multinational enterprises (MNEs) affected by cultural differences. Within this broad question, detailed issues need to be addressed, such as:  Are there lasting effects associated with countries of origin?

New Perspectives in IB Research

3

 How do cultural differences affect perceptions of relational risk?  What are the determinants of interpersonal and inter-unit trust? Linked to the increasing significant of culture is the recognized importance of institutions in the IB literature (Rugman & Brewer, 2001). Institutions represent the major immobile factors in a globalized market, especially when considered against the recent sensational mobility of firms and factors of production. In an international environment, legal, political and administrative systems tend to be the internationally immobile framework whose costs determine the international attractiveness of a location. A range of topics related to the governance of MNEs are flourishing. These topics refer to how to best govern parent–subsidiary relationships (Harzing, Sorge, & Paauwe, 2002) and explore the role of trust in the governance of inter-firm relationships (Williamson, 1975). Within this framework, the question arises of how to govern places for designing effective economic outcomes and how international competitive advantage is constructed (Audretsch & Feldman, 1996; Braunerhjelm & Feldman, 2006). Academics and practitioners interested in global strategy are also becoming more interested in better understanding global strategies related to the genesis of entrepreneurs (Oviatt & McDougall, 2005). Here the role of the international as well as of the social, ethical and environmental dimensions on capabilities creation become important (Zahra, 2005; Oviatt, McDougall, Simon, & Shrader, 1993) in order to answer to the broad question of what are the factors that give rise to international entrepreneurs. Scholars and practitioners will benefit from detailed case studies focusing on the role of local milieu, ethnic ties and R&D management in promoting the success of international entrepreneurs. Continuing globalization increases our need also to understand how financial relationships affect entrepreneurship and how these changes impact the various stakeholders of the firm (Helms, 2006). Systematic research on the myths about microfinance can therefore help in making progress in this direction. Similarly, the early emphasis on the transfer of MNE’s technology to host countries has recently shifted the literature towards a more comprehensive view of the links between internationalization and innovation. Attention has shifted to the process of ‘global knowledge creation and exchange’ which, calls for consideration of aspects like the nature of FDI decisions involving the international dispersion of innovative activities (Dunning, 1992; Buckley, Clegg, & Wang, 2003) and of international technology sourcing strategies (Dunning, 1988; Ahuja & Katila, 2001; Graebner, 2004). Therefore, a large interest in IB research is still devoted to assessing the

4

MARYANN FELDMAN AND GRAZIA D. SANTANGELO

role of knowledge in investment decisions (Singh, 2007). In particular, the following points are particularly salient:  Do localized knowledge spillovers complement knowledge asymmetries in international R&D investment decisions?  Do different levels of absorptive capacity of host firms call for different mechanisms through which spillover occurs?  What are the salient issues in firms acquisitions?

3. VOLUME ORGANIZATION The volume is organized in four main parts. The first addresses the management of cultural differences across countries, the second part deals with questions about the Governance of multinationals and places, the third part focuses on the genesis of international entrepreneurs and the fourth part on technology and international expansion. Part I, Managing Cultural Differences across Countries, has three chapters that address the lasting effects of country of origin influences on the transfer of organizational practices, the cultural perceptions of relational risk and the determinants of trust, at both the individual and organizational unit. In Chapter 2, Anne-Wil Harzing and Niels Noorderhaven examine headquarter–subsidiary relationship. Specifically they examine the effects of country origin on the transfer of management practices. The fact that multinational corporations are rooted in a cultural context that is specific to their countries of origin is often ignored. For example, US multinationals will have a different cultural orientation than UK, German or Japanese multinationals and this will affect their relationships with subsidiaries. Based on a survey of 150 MNE subsidiaries in four different countries over two time periods, they demonstrate that country of origin cultural differences have an enduring influence on the practices employed within these organizations. In particular, strong differences between US and Japanese MNEs endure over time. MNEs from the United States and from the United Kingdom, already very similar in 1995, have become even more alike in 2002. German MNEs demonstrate the most diversity: while remaining similar to Japanese MNEs over the study period, in some limited respects German MNEs seem to have converged towards US/UK practices. Chapter 3 contends that an important component of cultural differences relates to receptions of risk. In this chapter, Susana Costa e Silva and Luciara Nardon examine the role of cultural differences and perception

New Perspectives in IB Research

5

of relational risk in foreign partnerships. Based on an exploratory study of Portuguese managers doing business internationally, they find the unexpected result that cultural differences were perceived positively in certain circumstances and associated with lower relational risk. Based on this finding, they suggest that culture distance is an asymmetric construct in which the perception of a cultural difference may be interpreted as positive or negative depending on the perspective taken and the nature of the task at hand. Their results caution that the application of cultural distance indexes should take into account these contradictory effects when investigating managerial behavior. Chapter 4 by Kristiina Ma¨kela¨, Wilhelm BarnerRasmussen and Ingmar Bjo¨rkman examines trust in multinational corporations, examining the drivers at both interpersonal and inter-unit levels. Trust is conceptualized as a manifestation of the relational dimension of social capital, between interaction partners in multinational corporations at interpersonal and inter-unit levels of analysis. Their study is based on two quantitative datasets from the Finnish subsidiaries of foreign MNEs, 265 individual records and 102 unit level observations. Their results indicate that the drivers of trust exhibit similar patterns across both levels of analysis, but are stronger at the interpersonal level. Trust was significantly and positively related to the length of the relationship between the two individuals or units, and to the frequency of the communication between them whereas it was found to be unrelated with cultural distance. For practicing managers, the main message is that communication frequency and the length of the relationship matter for the relational social capital that exists within MNEs, at both at the interpersonal and inter-unit levels. Governance is the focus of Part II of the volume. An actor-centred conflict perspective is proposed in Chapter 5 by Susanne Blazejewski, who accounts for corporate coordination instruments as well as individual actor’s reactions on the local level of the MNE subsidiaries. By means of a qualitative empirical study on cultural control mechanisms applied by a German MNE to its Japanese subsidiary, Susanne Blazejewski is able to trace local conflicts and the Japanese actors’ micro-political responses arising during the implementation of the so-called cultural core values developed by headquarters. The governance of firms market relationships is investigated in Chapter 6 by Diana Weinberg and Abraham Carmeli, who explore how high-quality relationships, manifested by trust, respectful engagement and vitality, augment relationship commitment between importer and exporter. Although examples of governance abound in the study of export theory, scant attention has been devoted to dyadic relationships (importers and exporters) in import theory in the IB literature.

6

MARYANN FELDMAN AND GRAZIA D. SANTANGELO

The analysis of the data collected from 97 importing companies show that both trust and respectful engagement had a positive effect on relationship commitment, while vitality mediated the relationship between respectful engagement and relationship commitment. The presence of product substitutes seem also to have a significant impact on relationship commitment. Chapter 7 by Maryann Feldman and Roger Martin is a reprint from an article published in Research Policy, where the authors aim to advance economic development theory through the concept of jurisdictional advantage by demonstrating how places might strategically position themselves to gain economic advantage and then considering how this placespecific advantage might be constructed. The term ‘jurisdiction’ is used by Feldman and Martin to define the set of actors that have a common interest in a spatially bound community and therefore able to influence social and economic outcomes within their boundaries. Part III deals with the Genesis of International Entrepreneurs. In Chapter 8, Shameen Prashantham, Amer Qureshi and Stephen Young seek to extend understanding of the ‘international’ dimension of comparative international entrepreneurship within a global industry, namely the software industry. More specifically, they focus on two local ecologies, namely a regional agglomeration (Bangalore, India) and less developed (and known) niche (Lahore, Pakistan). Based on in-depth interviews in Bangalore and Lahore, exemplar case studies from both sub-national regions highlight the role of cross-border linkages between milieux illustrating that the global nature of the software industry and the central role of the innovative milieu in the USA have important implications for coordination and integration of the entrepreneurial processes of opportunity discovery, evaluation and exploitation across frontiers. While Indians have influenced the development of Silicon Valley, their ties with Bangalore seem primarily to be based on hard-nosed business relations. But in relation to Pakistan, while the US milieu is critical for all aspects of the entrepreneurial process, closed networks may be a barrier to long-term growth. In Chapter 9, Lisa Hewerdine and Catherine Welch re-examine the case of Cochlear (an Australian company that produces hearing implants for the deaf), challenging the classification of the company as ‘a prime example of a company that was ‘born global’. Although the company commenced international operations almost immediately upon establishment, their reanalysis of the Cochlear case suggests that to denote it a ‘born global’ or ‘international new venture’ is problematic, and a denial of the protracted process of evolution that took place. By contrast, Lisa Hewerdine and Catherine Welch contend that any account of Cochlear should not exclude or

New Perspectives in IB Research

7

downplay the period, prior to the existence of the company, in which the research and development of its first invention took place, or the role of the corporate parent and its organizational collaborators. In Chapter 10, Roy Mersland and R. Øystein Strøm examine myths about microfinance and conclude that microfinance – the provision of financial services to the poor – is a viable business model. Specifically, they evaluate three myths regarding microfinance based on new data from rated microfinance institutions. The first myth is that an efficient microfinance institution needs to be shareholder owned; second that its governance should first and foremost address the potential conflict between owners and managers; and third that microfinance institutions are drifting away from their poorer customers towards serving the wealthier. The data do not support any of these myths. The results point towards an industry finding its own viable business model, establishing its own way of providing banking services to a customer segment formerly unattached to the economy at large. The microfinance industry is showing that it is able to reach poor customers in a financially sustainable way. In Part IV, Technology and International Expansion are investigated. More specifically, the first two chapters deal with the issue of spillovers and absorptive capacity within the context of FDI inflows. In Chapter 11, Maria Luisa Petit, Francesca Sanna-Randaccio and Roberta Sestini analyze how firms’ R&D investment decisions are affected by asymmetries in knowledge transmission, taking into account different sources of asymmetry such as unequal know-how management capabilities and spillovers localization within an international oligopoly. Within an oligopoly model with endogenous R&D, they find that a better ability to manage knowledge flows incentivates the firm to invest more in R&D and that one-way FDI stimulates the MNE to raise its own R&D. Furthermore, it emerges that when geographical proximity increases the MNE’s capability to source local know-how, FDI is more likely to occur. In Chapter 12, Lamia Ben Hamida and Philippe Gugler examine intra-industry spillover effects from inward FDI in Swiss manufacturing firms by accounting for the mechanisms by which spillovers occur (viz. the increase of competition, demonstration effects and worker mobility), and the interaction between such mechanisms and the levels of domestic absorptive capacity. The results obtained suggest that high technology firms benefit from FDI heightening competition, while mid-technology firms benefit from demonstration effects. Low technology firms, which are not able to benefit from foreign affiliates via demonstration effects alone, manage to reap the benefit via the recruitment of MNEs labor. In addition, only firms which largely invest in absorbing foreign technology

8

MARYANN FELDMAN AND GRAZIA D. SANTANGELO

benefit from spillovers. Motivations of external corporate expansion are then addressed by Christoph Grimpe and Katrin Hussinger in Chapter 13, where they show whether and how the importance of different takeover motivations changes along the deal value distribution. Based on a comprehensive dataset of 652 European mergers and acquisitions occurred between 1997 and 2003, the results of the quartile regressions indicate that the importance of technological assets is higher for smaller target firms, which seem to complement the acquirer’s technology portfolio, while the importance of non-technological assets seems to be higher for larger targets, which tend to be used to gain access to international markets. The implications drawn by the authors point out that managers in the acquiring firm should be aware that they might overpay for the technological assets of a small firm, while they should carefully tailoring a well-developed integration strategy when acquiring larger targets.

REFERENCES Ahuja, G., & Katila, R. (2001). Technological acquisitions and the innovation performance of acquiring firms: A longitudinal study. Strategic Management Journal, 22(3), 197–220. Audretsch, D. B., & Feldman, M. P. (1996). R&D spillovers and the geography of innovation and production. American Economic Review, 86, 630–640. Bolino, M. C., Turnley, W. H., & Bloodgood, J. M. (2002). Citizenship behaviour and the creation of social capital in organizations. Academy of Management Review, 27(4), 505–522. Braunerhjelm, P., & Feldman, M. P. (Eds). (2006). Cluster genesis: The origins and emergence of technology-based economic development. Oxford: Oxford University Press. Buckley, P. J., Clegg, J., & Wang, C. (2003). Is the relationship between inward FDI and spillover effects linear? An empirical examination of the case of China. Presented at the 29th Annual Conference of European Academy of International Business, Copenhagen. Dunning, J. H. (1988). The eclectic paradigm of international production: A restatement and some possible extensions. Journal of International Business Studies, 19(1), 1–31. Dunning, J. H. (1992). Multinational enterprises and the global economy. Wokingham, England: Addison-Wesley. Ferner, A. (1997). Country of origin effects and human resource management in multinational companies. Human Resource Management Journal, 7, 19–37. Graebner, M. (2004). Momentum and serendipity: How acquired leaders create value in the integration of technology firms. Strategic Management Journal, 25, 751–777. Harzing, A.-W., Sorge, A., & Paauwe, J. (2002). Headquarters–subsidiary relationships in multinational companies: A British–German comparison. In: M. Geppert, D. Matten & K. Williams (Eds), Challenges for European management in a global context – Experiences from Britain and Germany. Houndmills, Basingstoke: Palgrave. Harzing, A. W. K., & Sorge, A. M. (2003). The relative impact of country-of-origin and universal contingencies on internationalization strategies and corporate control in

New Perspectives in IB Research

9

multinational enterprises: World-wide and European perspectives. Organisation Studies, 24(2), 187–214. Helms, B. (2006). Access for all. Washington, DC: CGAP. Hofstede, G. (1980). Culture’s consequences: International differences in work-related values. Beverly Hills, CA: Sage. Hofstede, G. (1991). Cultures and organizations: Software of the mind. London: McGraw-Hill. Kluckhohn, F. (1954). Culture and behavior. In: G. Lindzey (Ed.), Handbook of social psychology (Vol. 2, pp. 921–976). Cambridge, MA: Addison-Wesley. Kostova, T., & Roth, K. (2002). Adoption of an organizational practice by subsidiaries of multinational corporations: Institutional and relational effects. Academy of Management Journal, 45(1), 215–233. Leung, K., Bhagat, R. S., Buchan, N. R., Erez, M., & Gibson, C. B. (2005). Culture and international business: Recent advances and their implications for future research. Journal of International Business Studies, 36, 357–378. Noorderhaven, N. G., & Harzing, A. W. (2003). The ‘‘country-of-origin effect’’ in multinational corporations: Sources, mechanisms and moderating conditions. Management International Review, 43(Special issue 2), 47–66. Oviatt, B. M., & McDougall, P. P. (2005). Defining international entrepreneurship and modelling the speed of internationalization. Entrepreneurship Theory and Practice, 29, 537–554. Oviatt, B. M., McDougall, P. P., Simon, M., & Shrader, R. C. (1993). Heartware International Corporation: A medical equipment company born international, Part A. Entrepreneurship Theory and Practice, 18(2), 111–128. Rugman, A. M., & Brewer, T. L. (Eds). (2001). Oxford handbook of international business. Oxford: Oxford University Press. Singh, J. (2007). Asymmetry of knowledge spillovers between MNEs and host country firms. Journal of International Business Studies, 38(2), 764–786. Williamson, O. E. (1975). Markets and hierarchies, analysis and antitrust implications: A study in the economics of internal organization. New York: Free Press. Zahra, S. A. (2005). A theory of international new ventures: A decade of research. Journal of International Business Studies, 36, 20–28.

This page intentionally left blank

PART I MANAGING CULTURAL DIFFERENCES ACROSS COUNTRIES

This page intentionally left blank

HEADQUARTERS–SUBSIDIARY RELATIONSHIPS AND THE COUNTRY-OF-ORIGIN EFFECT Anne-Wil Harzing and Niels Noorderhaven ABSTRACT Purpose – Multinational corporations (MNCs) are often presented as harbingers of global practices, thus promoting a process of convergence between different national business systems. However, this view disregards the fact that MNCs, too, are rooted in countries of origin, and that this may have an enduring influence on the practices employed within these organizations. This chapter aims to throw new light upon this issue. Methodology – Using a survey methodology we compare company practices for around 150 MNC subsidiaries from four different countries (the USA, the UK, Germany, and Japan) at two points of time (1995 and 2002). Findings – Our findings show that even the most internationalized companies in the world continue to show unique country patterns. In particular, strong differences between US and Japanese MNCs endure. However, MNCs from the US and from the UK, already very similar in 1995, have become even more alike in 2002. German MNCs show an interesting picture. While remaining similar to Japanese MNCs (and very

New Perspectives in International Business Research Progress in International Business Research, Volume 3, 13–40 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1745-8862/doi:10.1016/S1745-8862(08)03002-1

13

14

ANNE-WIL HARZING AND NIELS NOORDERHAVEN

different from their US and UK counterparts), in some limited respects German MNCs seem to have adapted to the US/UK practices. Originality – Previous studies, focusing at different groups of countries and different practices, and conducted at various points in time, have been insufficiently cumulative to form a basis for firm conclusions. This study provides a systematic comparison of MNC company practices at two points in time.

INTRODUCTION In the globalization debate the multinational corporation (MNC) is often presented as a harbinger of global practices (Dicken, 1998). However, even though business may become increasingly global in many respects, the MNC remains dependent upon certain local environments for its competitive advantages and renewal thereof (So¨lvell & Zander, 1995). Moreover, far from being ‘nationless’ organizations, as suggested by Ohmae (1990), even the most global MNCs in many respects still appear to be strongly rooted in their country-of-origin (Hu, 1992; Ruigrok & Van Tulder, 1995). A small but growing body of literature discusses this ‘country-oforigin’ effect in MNCs (for overviews, see Ferner, 1997; Harzing & Sorge, 2003; Noorderhaven & Harzing, 2003). Pauly and Reich (1997), looking at MNCs from the US, Japan, and Germany, conclude that the behavior of the firms studied divides into three distinct ‘syndromes’, typical of the respective national origins; and that these ‘syndromes are durably nested in broader domestic institutional and ideological structures’ (Pauly & Reich, 1997, p. 24). Ngo, Turban, Lau, and Lui (1998) studied the effect of the nationality of the parent company on human resource (HR) practices of subsidiaries operating in Hong Kong. Comparing these practices for subsidiaries with parent firms from the US, Great Britain, Japan, and Hong Kong itself, they find strong support for the hypothesis that countryof-origin influences the firms’ human resource management (HRM) practices. Lubatkin, Calori, Very, and Veiga (1998) focus on the administrative approach used by headquarters (HQ) in recently acquired subsidiaries in Britain and France. During the transition period following an acquisition, the initial control strategies employed by the parent firm are seen as reflecting the acquiring firm’s beliefs about ‘how things ought to be done’ (Lubatkin et al., 1998, p. 671). They conclude that British and

Headquarters–Subsidiary Relationships and Country-of-Origin Effect

15

French parent firms tend to establish different HQ–subsidiary relationships. Most recently, based on data on 287 subsidiaries from 104 MNCs, Harzing and Sorge (2003) conclude that country-of-origin comes forward as one of the most important predictors of the control mechanisms used by MNCs, while also influencing their overall internationalization strategy to some extent. However, not all the evidence points in the same direction. Tregaskis (1998) conducted an analysis comparable to that of Ngo et al. (1998) for firms operating in Britain, comparing nationally owned companies with subsidiaries of MNCs from continental Europe, the US, and Japan. But in contrast with Ngo et al. (1998), she found only limited differences in HR development practices associated with the parent company’s national origin. Likewise, Lindholm (1999–2000) found that the European MNC he studied adopted standardized performance management policies and practices both in its home country and in overseas subsidiaries, and that these policies and practices had a broadly similar impact on the job satisfaction of hostcountry employees in different subsidiaries. Hayden and Edwards (2001), although stating that ‘MNCs continue to be firmly embedded in, and strongly influenced by, their country of origin’ (p. 132), nevertheless observed that the country-of-origin effect in a large Swedish MNC eroded as foreign, mainly Anglo-American, practices were adopted. In a comparison of HRM practices in subsidiaries US, Japanese, and German MNCs, Pudelko and Harzing (2007) find that Japanese and German subsidiaries are increasingly adopting US ‘‘best practices’’, while US subsidiaries themselves show a mix of country-of-origin and localization effects. With regard to German MNCs in particular, Lane (2000, 2001) suggests that they might have deviated from established societal patterns in the second part of the nineties and that their practices have become more similar to those of British and American MNCs. While all of these studies provide useful information, it is difficult to compare them systematically as they were conducted in different time periods and focus on very different organizational practices. This chapter therefore attempts to provide a comparison of country-of-origin effects for a very wide range of aspects of the HQ–subsidiary relationship and for two time periods: 1995 and 2002. We assume that if globalization does indeed bring about a process of convergence of cultural, political, and economic aspects of life (Giddens, 1999), this should be reflected in a weakening of the country-oforigin effect, and increasing homophily of MNCs from different national origins. We test this general conjecture by comparing differences between MNCs from the US, the UK, Germany, and Japan, in 1995 and in 2002.

16

ANNE-WIL HARZING AND NIELS NOORDERHAVEN

In the next sections we will first discuss our choice of organizational practices and countries. We then provide an overview of our methods and the results of our empirical investigation. A short discussion and conclusion section concludes the chapter.

CHOICE OF CONCEPTS AND COUNTRIES In this chapter we look at differences in the HQ–subsidiary relationship between MNCs from different countries of origin. This relationship can be seen as a classic control problem, whose attributes are similar to principal– agent relationships (Nohria & Ghoshal, 1994). HQ, the principal, cannot make all decisions because it does not possess all the necessary knowledge or resources, but it cannot leave all decisions to subsidiaries because the interests of subsidiaries might be different from that of HQ or the MNC as a whole. Therefore, the key aspect of the HQ–subsidiary relationship is the way in which HQ ensures that subsidiaries are working towards common organizational goals. The different types of control mechanisms are the tools that HQ have to achieve this alignment. Hence, the level of control exercised by HQ by means of the different types of control mechanisms is the first element of the HQ–subsidiary relationship that we will investigate. As we will see below, there is a range of control mechanisms available that goes beyond the level of autonomy granted to subsidiaries. The second element that we will look at is the level of expatriate presence in subsidiaries. Expatriates can perform many roles in the HQ–subsidiary relationship, among them control and knowledge transfer. The level of interdependence between HQ and subsidiaries in comparison to the level of interdependence between subsidiaries is a third important element of the HQ–subsidiary relationship. Another element that we will study is the level of local responsiveness – in terms of local production, local R&D, and adaptation of products and marketing to local conditions – that HQ allows to the subsidiary. On all these aspects we will compare data from a survey conducted in 1995 with data from a survey conducted in 2002. In addition, we will use 2002 data to look at country-of-origin effects for another aspect of the HQ relationship that has been almost neglected so far: the existence of language barriers between HQ and subsidiaries and the language policies applied by HQ. Finally, as performance is a major outcome variable in most management studies, we will also look for country-of-origin effects in relative performance.

Headquarters–Subsidiary Relationships and Country-of-Origin Effect

17

To ensure a large enough sample size for individual countries, we chose to focus our analysis on four MNC HQ countries: the US, the UK, Germany, and Japan. A focus on these four countries is very appropriate for several reasons. First, they are the four most highly ranked developed countries in terms of their GDP (World Bank, 2004). Second, they all host a very significant number of MNCs. And third the four countries differ significantly in both their culture and business system. Ronen and Shenkar (1985) discussed nine different studies that investigated cultural differences and identified clusters of countries. In seven of these studies the UK was included and in all studies it was classified in the Anglo cluster, together with the U.S. Although Germany and Japan show significant cultural differences and were always classified in different clusters, in some respects Germany takes a middle position between Japan and the UK and the US. In terms of Hofstede’s dimensions, Germany scores similar to the US and the UK on masculinity and power distance, but it scores mid-way between the US, the UK, and Japan on uncertainty avoidance and individualism. Looking at their business systems, the four countries also differ dramatically. Britain shares with the US an adherence to consumer capitalism, which is in strong contradiction to the producer capitalism more typical of both Germany and Japan. With the first comes a focus on marketing excellence, while the second is characterized by manufacturing excellence. There are also major differences between Germany/Japan and UK/USA with regard to capital structure and the importance of stock markets. British and American companies raise their funds mainly by selling stock (are equity-based), while German and Japanese companies are mainly credit-based (Prowse, 1994). These different capital structures are also reflected in different philosophies about the management of companies. While Anglo-American companies are mainly managed in the interest of shareholders and focus on the maximization of short-term profits, German and Japanese companies are more concerned about long-term viability and stability. This phenomenon is reinforced by the fact that in the AngloAmerican countries, around 80% or more of the shares are held for trading purposes, while in Germany and Japan the overwhelming majority of shares are held for control purposes (Prowse, 1994, p. 24, Table 3). Obviously, investors holding shares for trading purposes are more likely to focus on short-term returns than on long-term stability. The consequence of these differences might be that the product and its production and development are more important for the corporate identity of a multinational from Germany and Japan. As a result, it will be more likely to promote an international strategy in the process of going

18

ANNE-WIL HARZING AND NIELS NOORDERHAVEN

international. This means it will attempt to perform on the basis of an existing product template and its advantages, it will try to replicate this product template abroad and emphasize interdependencies or identity of the country-of-origin template and the subsidiary template, and it will not go for multi-domestic or other locally responsive strategies abroad. The British and US multinationals, however, will see the enterprise as hanging together around financial flows and measures and encourage marketing postures, which are more multi-domestic or locally responsive. In this way, internationalization strategies are likely to be the consequence of deeply rooted, societally embedded strategies in the country-of-origin. We would therefore expect MNCs from Germany/Japan and USA/Britain to differ in the way they internationalized and hence in their HQ–subsidiary relationships. Several studies have indicated that German MNCs display a smaller geographical reach than British MNCs, until recently relied more heavily on export than on FDI, and are deeply embedded into their domestic business system, producing far more value from their domestic base than in their foreign affiliates (Ruigrok & van Tulder, 1995; Do¨rre, 1996; Hirst & Thompson, 1996; Lane, 1998; Whitley, 1998). According to Lane (1998) the foreign affiliates of German MNCs are replicas of their parent company, rather than adapting to host-country features, while British MNCs tend to follow more of a conglomerate strategy with a lot of subsidiaries resulting from acquisitions. These differences would point into the direction of a lower local responsiveness and higher dependence on HQ for German subsidiaries, while the reverse would be true for British subsidiaries. Whitley (1999) discusses this even more directly. He describes German MNCs as co-operative hierarchies, in which most foreign subsidiaries of any significance will be quite closely supervised and integrated into parent activities and where the integration of foreign subsidiaries into host economies is limited. Whitley sees the isolated firm type as more typical of American and British companies. In this type of firm subsidiaries are managed at a distance and provided the formal procedures and targets are followed, units will be allowed some local adaptation and will not be as fully integrated into their parents’ operations as is the case with co-operative firms. The result may be more integration into host economies with local sourcing and adaptation of products to local markets. Subsidiaries from co-operative hierarchies will rely more on products and technologies from the parent. We could of course wonder to what extent country-of-origin effects are enduring and hence whether convergence might become more important

Headquarters–Subsidiary Relationships and Country-of-Origin Effect

19

in the 21st century. Germany in particular has recently experienced more rapid changes in ownership and governance moving more towards AngloAmerican models, although the question is to what extent conversion to promoting shareholder value is just lip service (Lane, 2000; Morgan & Kristensen, 2005). This might also impact on the way German MNCs structure their HQ–subsidiary relationships. Lane (2000, 2001) suggests that German multinationals might have deviated from established societal patterns in the second part of the nineties. She mentions that subsidiaries of German multinationals have been allocated more resources and granted more autonomy, with the organizations moving towards a decentralized network structure and subsidiaries becoming more embedded in the local environment, through outsourcing and local recruitment of managers. However, her study is based on public company documents and secondary data and covers only seven of the most internationalized German MNCs and she calls for ‘‘more and better data’’ (Lane, 2001, p. 93). Our study uses primary data collected at some 30–50 subsidiaries from MNCs headquarted in four different countries and compares two distinct time periods. By comparing MNCs from the same countries across two different time periods, we should be able to give a more comprehensive picture of possible changes in country-of-origin effects.

METHODS Data were collected by means of two large-scale international mail surveys. In both cases the questionnaire was developed after an extensive review of the relevant literature on HQ–subsidiary relationships and was pilot-tested with academics in the area, managers, and international students. The first study was conduced in 1995/1996 by mailing questionnaires to the managing directors of wholly owned subsidiaries of 122 multinationals in 22 different countries, representing 8 manufacturing industries. The second study was conducted in 2002 by mailing questionnaires to the managing directors of wholly owned subsidiaries of 82 MNCs headquartered in the USA, Japan, Germany, the UK, France, and the Netherlands. Subsidiaries were located in more than 50 different countries and 4 very different manufacturing industries were represented. Disregarding undeliverable questionnaires, the overall response rate in the first study was 20%. The total number of 287 subsidiary responses represented 104 different HQ (85% of our population). In the second study the response rate was 8%. The resulting

20

ANNE-WIL HARZING AND NIELS NOORDERHAVEN

sample of 169 subsidiaries represented nearly 50 different MNCs (61% of our population). Although the response rate in the second study was much lower than for the first study, it is not unusual for multi-country studies with high-level executives as respondents. Harzing (1997) reported that response rates for international mail surveys typically varied between 6% and 16% and key studies in the field (e.g. Ghoshal & Nohria, 1989) have been based on response rates of 15%. Ghoshal and Nohria’s data were collected nearly twenty years ago. Intensification of the pace of business as well as the increasing use of mail surveys are likely to have led to a substantial decline in willingness to respond to mail surveys. In this chapter we only use the data for subsidiaries of MNCs headquartered in the USA, Japan, Germany, or the UK, which resulted in a sample size of 149 for 1995 and 145 for 2002. Non-response bias was evaluated in a number of ways. First, we tested whether responses on the key variables in this study differed systematically between respondents in the original mailing and respondents in the reminder. In this procedure late respondents are treated as a proxy for non-respondents. No significant differences were found for any of the key variables in our study. Second, we compared responding and nonresponding firms on size (number of employees), age, industry and country of HQ. No significant differences were found on any of the variables. We can therefore be reasonably confident that non-response bias is not a problem in our study. Measures for the key concepts in our study were based on a combination of existing and newly created scales. Full details can be found in Appendix 1. Our surveys used a key-informant approach and our results are therefore based on the information of a single respondent in each organization. This is a limitation that this study shares with virtually all large-scale studies of multinationals. The prevalent response rates in international mail surveys make another approach practically infeasible. Second, although every care was taken to formulate questions as unambiguously as possible, our study used perceptual measures to operationalize some of the constructs. This was done first because of the not immediately quantifiable nature of concepts such as control mechanisms. The result is that the answers to our questions might contain an element of perception, which might reduce the validity of our findings. However, questions elicited information on actual practices and policies, rather than opinions on such practices that might be personally colored and depend on the person of the respondent instead of on the organization.

21

Headquarters–Subsidiary Relationships and Country-of-Origin Effect

RESULTS Table 1 compares the sample for the four different countries on basic descriptives. Our sample sizes are very similar for the two periods and although for the individual countries they are not very large, they are large enough for basic statistical comparisons. On average subsidiaries from Japanese MNCs are clearly younger than subsidiaries from other MNCs, reflecting Japan’s shorter history of internationalization. Although in 1995 German and US subsidiaries are substantially larger than British and Japanese subsidiaries, these differences are not significant and mostly caused by a small number of outliers. In 2002 our subsidiaries are very similar in size.

Control Mechanisms Table 2 first looks at the type of control mechanisms that HQs in the four different countries apply towards their subsidiaries. Our 1995 data show that impersonal control is the most favored means of control in subsidiaries of all four countries, while direct personal control (centralization and direct supervision) is the least favored means of control. There are, however, some differences between countries, with Japanese subsidiaries showing particularly low level of impersonal control mechanisms and also slightly lower levels of control by socialization and networks (shared values, international training and task-forces, and informal communication), while both Japanese and German MNCs show higher levels of direct personal control. The 2002 Table 1.

Overview of Differences in Descriptives in 1995 and 2002.

Variable

Median Scores for Different Aspects of the HQ–Subsidiary Relationship in British, German, Japanese, and US MNCs British MNCs German MNCs

Descriptives 1995 Sample size Subsidiary age (years) Subsidiary size (employees) Descriptives 2002 Sample size Subsidiary age (years) Subsidiary size (employees)

Japanese MNCs

US MNCs

25 37 170

32 36 450

38 18 178

54 33 400

30 42 131

36 36 153

28 20 185

51 36 110

22

Table 2. Variable

Overview of Differences in Control Mechanism in 1995 and 2002.

Mean Scores for Different Aspects of the HQ–Subsidiary Relationship in British, German, Japanese, and US MNCs

Control mechanisms 1995 data Direct personal 3.27 control (1–7) Impersonal control 5.27 (1–7) Indirect personal 4.53 control (1–7) Significant differences Control mechanisms 2002 data Centralization 2.29 upstream (1–5) Centralization 2.24 downstream (1–5) Impersonal control 4.83 (1–7) Indirect personal 4.38 control (1–7) Significant differences

Difference between German and Japanese MNCs

Difference between British and US MNCs

Difference between Japanese and US MNCs

German MNCs

Japanese MNCs

US MNCs

Sign. level

Sign. at .05/.10?

Sign. level

Sign. at .05/.10?

Sign. level

Sign. at .05/.10?

Sign. level

Sign. at .05/.10?

3.69

3.48

2.99

.221

No

.516

No

.353

No

.069

(Yes)

5.00

4.27

5.13

.311

No

.004

Yes

.825

No

.000

Yes

4.53

4.26

4.77

.996

No

.280

No

.529

No

.021

Yes

0/3

1/3

0/3

3/3

2.57

3.13

2.12

.222

No

.032

Yes

.373

No

.000

Yes

2.11

2.06

2.13

.429

No

.763

No

.559

No

.710

No

5.07

4.08

5.09

.350

No

.001

Yes

.299

No

.000

Yes

4.19

3.75

4.36

.394

No

.068

(Yes)

.935

No

.021

Yes

0/4

3/4

0/4

3/4

ANNE-WIL HARZING AND NIELS NOORDERHAVEN

British MNCs

Difference between British and German MNCs

Headquarters–Subsidiary Relationships and Country-of-Origin Effect

23

data confirm this picture with Japanese subsidiaries again showing lower levels of impersonal control and indirect personal control. The 2002 data, however, measured centralization (one aspect of direct personal control) in a more direct way by asking subsidiaries about their level of autonomy in a range of areas. Although MNCs in the four countries do not differ in the extent of autonomy they grant their subsidiaries for downstream activities (marketing, finance, and HRM), Japanese subsidiaries have a significantly lower level of autonomy than subsidiaries from all other countries for upstream activities (product development and modification, selection of and negotiation with suppliers). Both Japanese and German MNCs seem to differentiate the level of autonomy they grant to subsidiaries depending on the type of activities (less for upstream than for downstream), while for British and US MNCs the level of autonomy for both type of activities is virtually identical.

Expatriate Presence One aspect of control is the presence of expatriates in subsidiaries. Table 3 shows that in 1995 there was a clear distinction between German and Japanese subsidiaries on the one hand and British and US subsidiaries on the other. The former had a higher proportion of expatriates in top-5 functions and in particular had an expatriate as managing director in more than half of the subsidiaries. Our 2002 data show that this picture has not changed at all for Japanese subsidiaries. The percentage of expatriate managing directors is virtually identical to that in 1995. And although the percentage of expatriates in top functions is lower, the number of functions included has been expanded to 10 and hence the actual number of expatriates in top functions is virtually identical to the 1995 numbers as well. In contrast, German MNCs seem to have localized considerably in the timespan of 7 years, reducing the number of expatriates employed in subsidiaries, to the extent that these numbers now resemble those of British and US MNCs rather than Japanese MNCs.

Functions of Expatriation In terms of the functions that expatriates fulfil within the subsidiaries, there are some substantive differences between countries. Starting with the functions related to control, our 1995 data show that expatriates in

24

Table 3. Variable

Overview of Differences in the Presence and Function of Expatriates in 1995 and 2002. Mean Scores for Different Aspects of the HQ–Subsidiary Relationship in British, German, Japanese, and US MNCs

Expatriate presence 1995 data Managing director 27% PCN Number of expatriates 1.00 in top-5 (0–5) Ensuring HQ policies 2.08 are implemented (1–5) Improving 3.38 communication to HQ and subsidiaries (1–5) Ensuring 3.06 homogeneous corp. culture (1–5) Knowledge transfer 3.25 from HQ (1–5) Training for future 2.25 positions at HQ (1–5) Significant differences Expatriate presence 2002 data Managing director 17% PCN

Difference between British and US MNCs

Difference between Japanese and US MNCs

German MNCs

Japanese MNCs

US MNCs

Sign. level

Sign. at .05/.10?

Sign. level

Sign. at .05/.10?

Sign. level

Sign. at .05/.10?

Sign. level

Sign. at .05/.10?

56%

58%

19%

.049

Yes

.827

No

.453

No

.000

Yes

1.66

2.16

0.98

.034

Yes

.156

No

.944

No

.000

Yes

2.94

3.62

1.89

.056

(Yes)

.139

No

.609

No

.000

Yes

3.43

3.97

2.49

.894

No

.048

Yes

.026

Yes

.000

Yes

2.68

2.42

2.49

.299

No

.343

No

.140

No

.815

No

2.93

3.71

3.23

.292

No

.009

Yes

.950

No

.110

No

3.00

2.32

2.71

.057

(Yes)

.021

Yes

.211

No

.155

No

4/7 22%

54%

14%

.579

No

3/7 .009

Yes

1/7 .723

No

4/7 .000

Yes

ANNE-WIL HARZING AND NIELS NOORDERHAVEN

British MNCs

Difference between Difference between British and German German and MNCs Japanese MNCs

12%

12%

22%

6%

.869

No

.036

Yes

.110

No

.000

Yes

4.05

4.19

4.91

3.94

.799

No

.178

No

.830

No

.045

Yes

4.75

4.65

5.73

4.34

.866

No

.056

(Yes)

.420

No

.006

Yes

4.15

3.23

3.57

4.06

.091

(Yes)

.525

No

.862

No

.352

No

4.00

3.70

4.27

4.11

.572

No

.279

No

.820

No

.739

No

5.29

4.85

5.43

4.51

.393

No

.246

No

.110

No

.047

Yes

4.71

3.69

3.57

4.24

.040

Yes

.810

No

.318

No

.170

No

4.84

4.80

4.42

4.53

.920

No

.412

No

.489

No

.821

No

4.58

4.00

3.58

4.53

.294

No

.437

No

.925

No

.056

(Yes)

2/10

3/10

0/10

6/10

Headquarters–Subsidiary Relationships and Country-of-Origin Effect

Percentage of expatriates in top functions Ensuring HQ policies are implemented (1–7) Improving communication to HQ (1–7) Improving communication to other subsidiaries (1–7) Ensuring homogeneous corp. culture (1–7) Knowledge transfer from HQ (1–7) Knowledge transfer from other subsidiaries (1–7) Training for future positions at HQ (1–7) Training for future positions at other subsidiaries (1–7) Significant differences

25

26

ANNE-WIL HARZING AND NIELS NOORDERHAVEN

subsidiaries of Japanese MNCs are more likely to fulfil the ‘‘bear’’ role of expatriation (i.e. ensuring HQ policies are implemented), while to a lesser extent this is also true for German MNCs (see also Harzing, 2001). In 2002 this is still true for Japanese MNCs, but again German MNCs seem to have moved to Anglo-American practices in this respect. The most important function of expatriates in Japanese MNCs is improving communication to HQs (the ‘‘spider’’ role) and this function is more important for Japanese MNCs than for MNCs from any of the other three countries, but most particularly US MNCs. This is true in both 1995 and 2002. Our 2002 data show that it is communication to HQs that is most important for Japanese MNCs as the use of expatriates to improve communication to other subsidiaries is much less important. The same is true for German MNCs, while for British and US MNCs communication to HQs and subsidiaries seems to be almost equally important. Japanese and German MNCs seem to be more HQ-centred, while for British and US MNCs there is considerable interaction between subsidiaries. No significant differences are apparent between countries for the ‘‘bumble-bee’’ role of international transfers (i.e. ensuring a homogenous corporate culture), neither for 1995 nor for 2002. Knowledge transfer from HQ seems to be one of the most important functions of expatriates. In all four countries and for both time periods, it is indicated to be the first or second most important function of expatriation. As a result there are few significant differences between countries. Our 2002 data, however, included information on both knowledge transfer from HQ and knowledge transfer from other subsidiaries. And consistent with the observation above for British and US MNCs the knowledge function from subsidiaries is nearly as important as the knowledge transfer function from HQ, while for Japanese and German MNCs knowledge transfer from subsidiaries is much less important. Using expatriation as management development seems to be less important for Japanese MNCs than for MNCs from the other countries. This is true for both 1995 and 2002. Again, though, the 2002 data show the more HQ-centric nature of Japanese and German MNCs, with training for future positions at HQ to be more important than training for future positions at other subsidiaries, while for British and US MNCs these two functions are not substantively different.

Intra-Company Purchases and Knowledge Flows Table 4 provides an overview of the extent of interdependence in the subsidiaries in our two samples, measured in terms of their level of intra-company

Variable

Overview of Differences in Interdependence in 1995 and 2002.

Mean Scores for Different Aspects of the HQ–Subsidiary Relationship in British, German, Japanese, and US MNCs British MNCs

Interdependence 1995 data Purchases from HQ 5% Purchases from 32% subsidiaries Purchases from 63% external suppliers (estimate) Significant differences Interdependence 2002 data Purchases from HQ/ 22% subsidiaries in HQ country Purchases from other 24% subsidiaries Purchases from 54% external suppliers Knowledge flows with 3.34 HQ (1–7) Knowledge flows with 3.22 subsidiaries (1–7) Significant differences

Difference between Difference between Difference between Difference British and German and British and US between Japanese German MNCs Japanese MNCs MNCs and US MNCs

German MNCs

Japanese MNCs

US MNCs

Sign. level

Sign. at .05/.10?

Sign. level

Sign. at .05/.10?

Sign. level

Sign. at .05/.10?

Sign. level

Sign. at .05/.10?

58% 14%

51% 24%

14% 41%

.000 .005

Yes Yes

.365 .072

No (Yes)

.084 .311

(Yes) No

.000 .013

Yes Yes

28%

25%

45%

–

–

–

–

–

–

–

–

2/2

1/2

1/2

2/2

39%

44%

16%

.050

Yes

.601

No

.432

No

.000

Yes

18%

25%

28%

.734

No

.398

No

.500

No

.892

No

41%

31%

57%

.078

(Yes)

.253

No

.983

No

.005

Yes

3.55

3.39

3.39

.484

No

.585

No

.845

No

.974

No

2.83

2.78

3.18

.158

No

.853

No

.903

No

.059

(Yes)

2/5

0/5

0/5

Headquarters–Subsidiary Relationships and Country-of-Origin Effect

Table 4.

3/5

27

28

ANNE-WIL HARZING AND NIELS NOORDERHAVEN

purchases and knowledge flows. In 1995, German and Japanese subsidiaries were clearly different from British and American subsidiaries. The former showed a significantly higher dependence on HQ, while for the latter the dominant trading partners were other subsidiaries and external suppliers. While subsidiaries of German and Japanese MNCs tended to function mostly as pipelines for their HQs, subsidiaries of British and American MNCs could be seen as important nodes in the corporate network. In 2002, both German and Japanese subsidiaries seem to have moved away slightly from the dominant orientation to HQ, although this move was more pronounced for German subsidiaries than for Japanese subsidiaries. However, the main patterns are still similar: Japanese and German subsidiaries are more dependent on HQ and less dependent on external suppliers than British and American subsidiaries. We should note here that though the data for 1995 are only a rough estimate of actual intra-company purchases as we only asked respondents to tick rather broad categories that were subsequently converted into percentages by taking the mid-point of each category. Hence data for 2002, where we asked for the actual percentage, are likely to be more reliable. In the context of the continuing differences between German and Japanese MNCs on the one hand and British and American MNCs on the other hand, it is also illuminative to look at the median values for purchases from HQ and from external suppliers. For German and Japanese MNCs these are 30% and 17.5/5%, respectively, indicating a clear dominance of intra-company purchases. For British and American MNCs, the median value for purchases from HQ is 0%, while for external purchases it is 35% indicating a very clear dominance of external suppliers. Japanese and German MNCs are more tightly integrated, conducting core functions within the corporate network, while British and American MNCs have outsourced those functions to a larger extent. In terms of knowledge flows differences between the four countries are generally small. However, while for British and American MNCs flows with HQ and subsidiaries are of nearly equal importance, for Japanese and German MNCs flows with HQ are more important than flows with other subsidiaries, which confirms the more HQ-centric nature of MNCs from the latter countries.

Local Responsiveness While the level of intra-company sales measures the extent to which subsidiaries are integrated with the rest of the MNC network, Table 5 shows the extent to which they are locally responsive. Our data for 1995 show that,

Variable

Overview of Differences in Responsiveness in 1995 and 2002.

Mean Scores for Different Aspects of the HQ–Subsidiary Relationship in British, German, Japanese, and US MNCs

Local responsiveness 1995 Product modification Marketing modification Local manufacturing Local R&D Significant differences Local responsiveness 2002 Product modification (1–7) Marketing modification (1–7) Local manufacturing Local R&D Collaboration with local suppliers (1–7) Collaboration with local customers (1–7) Significant differences

Difference between Difference between British and German German and MNCs Japanese MNCs

Difference between British and US MNCs

Difference between Japanese and US MNCs

British MNCs

German MNCs

Japanese MNCs

US MNCs

Sign. level

Sign. at .05/.10?

Sign. level

Sign. at .05/.10?

Sign. level

Sign. at .05/.10?

Sign. level

Sign. at .05/.10?

data 44% 55%

15% 51%

26% 46%

31% 47%

.000 .711

Yes No

.040 .583

Yes No

.156 .416

No No

.457 .869

No No

65% 26%

29% 13%

38% 15%

43% 20%

.000 .024

Yes Yes 3/4

.253 .584

No No 1/4

.020 .380

Yes No 1/4

.591 .315

No No 0/4

data 4.33

3.50

3.21

4.04

.097

(Yes)

.544

No

.521

No

.064

(Yes)

5.13

4.53

5.17

5.16

.158

No

.172

No

.950

No

.958

No

45% 39% 4.07

46% 27% 3.22

25% 19% 3.29

49% 30% 3.51

.944 .189 .052

No No (Yes)

.047 .402 .890

Yes No No

.665 .319 .200

No No No

.015 .199 .626

Yes No No

5.53

4.19

4.54

4.79

.005

Yes

.521

No

.050

Yes

.590

No

3/6

1/6

1/6

Headquarters–Subsidiary Relationships and Country-of-Origin Effect

Table 5.

2/6

29

30

ANNE-WIL HARZING AND NIELS NOORDERHAVEN

overall, marketing shows the highest level of local responsiveness with around half of the marketing being adapted to local circumstances. Product modification occurs mainly within British subsidiaries, while these subsidiaries also show the highest proportion of local manufacturing and local R&D. Overall, localization of manufacturing is more common than localization of R&D. German subsidiaries in particular show a rather low level of local responsiveness. We should note here that the data for 1995 are only a rough estimate of actual level of local responsiveness as we only asked respondents to tick rather broad categories, which were subsequently converted into percentages by taking the mid-point of each category. Hence these data might not be completely accurate and can only be compared with 2002 on a general level. Our data for 2002 again show that marketing modification is more common than product modification and that local manufacturing is more likely than local R&D. British subsidiaries remain the ones that are most locally responsive, including with respect to two new measures of local responsiveness: collaboration with local suppliers and customers. However, in comparison to 1995, German subsidiaries seem to have increased their level of local manufacturing and R&D to such an extent that it is now no longer significantly different from British and American subsidiaries. In 2002, it is Japanese subsidiaries that show the lowest level of local responsiveness.

Language Barrier One aspect of the HQ–subsidiary relationship that has received very little attention so far is the existence of a language barrier between HQs and subsidiaries and policies that are in operation to overcome this barrier (see also Feely & Harzing, 2003; Harzing & Feely, 2008). This variable was not included in our 1995 survey and hence we can only discuss the most recent data. As Table 6 shows German and Japanese MNCs are far more likely to experience a language difference between HQ and subsidiaries (i.e. HQs and subsidiary managers having a different native language) than British and US MNCs. Of course this is not surprising given the larger number of countries that have English as their native language and the tendency of US and British MNCs to direct a large part of their foreign direct investment to countries that have English as their native language. And given the widespread use of English as a second language it is also to be expected that the capability of subsidiary staff in the HQ language for British and US MNCs is very high (around 6 on a 7-point scale). Conversely, the capability of subsidiary managers in Japanese and German is rather low (2 and 3.6 on a 7-point scale).

Variable

Language 2002 data % of subsidiaries with language difference Capability subsidiaries staff in HQ language (1–7) Corporate language ¼ HQ language Corporate language ¼ other language No official corporate language Capability subsidiaries staff in corp. language (1–7) % functions with cross-lingual communication Significant differences

Overview of language policies and barriers in 2002.

Mean Scores for Different Aspects of the HQ–Subsidiary Relationship in British, German, Japanese, and US MNCs

Difference between Difference between Difference between Difference British and German and British and US between Japanese German MNCs Japanese MNCs MNCs and US MNCs

British MNCs

German MNCs

Japanese MNCs

US MNCs

Sign. level

Sign. at .05/.10?

Sign. level

Sign. at .05/.10?

Sign. level

Sign. at .05/.10?

Sign. level

Sign. at .05/.10?

46%

97%

93%

41%

.000

Yes

.416

No

.637

No

.000

Yes

5.79

3.60

2.00

6.27

.000

Yes

.003

Yes

.117

No

.000

Yes

77%

14%

28%

80%

.000

Yes

.587

No

.695

No

.000

Yes

0%

66%

44%

0%

.000

Yes

.257

No

.756

No

.000

Yes

23%

20%

28%

20%

.749

No

.051

(Yes)

.705

No

.526

No

5.67

4.79

5.06

6.19

.075

(Yes)

.680

No

.115

No

.017

Yes

55%

59%

58%

51%

.668

No

.826

No

.607

No

.432

No

4/7

2/7

0/7

Headquarters–Subsidiary Relationships and Country-of-Origin Effect

Table 6.

5/7

31

32

ANNE-WIL HARZING AND NIELS NOORDERHAVEN

As a result, while British and US MNCs select their HQ language as the corporate language in all cases, German and Japanese MNCs can usually not afford to do so. Even so, 28% of the Japanese MNCs still insist on Japanese as their corporate language, while another 28% do not have a corporate language at all. As a result only 44% of the Japanese MNCs has selected another language (i.e. English) as their corporate language, while this is the case for 66% of the German MNCs. It is hence not surprising that one of the key functions of expatriates in Japanese MNCs is to improve communication to HQ. The capability of subsidiary staff in the corporate language in German and Japanese MNCs is lower than that of subsidiary staff in British and US MNCs (around 5 on a 7-point scale). However, we should not forget that the former includes subsidiaries where the corporate language is German or Japanese rather than English. The lower level of language difference between HQ and subsidiaries for British and US MNCs and the widespread use of English as a corporate language does not mean that British and US MNCs do not have to cope with language differences at all. In fact, when asked which functions, out of a list of eight, were involved in cross-lingual communications, British and US MNCs listed nearly as many as German and Japanese MNCs. Hence it appears that the role of language in HQ–subsidiary relationships is well worth further study.

Performance Differences A final comparison looks at performance differences between subsidiaries (Table 7). This variable was not included in our 1995 survey and hence we can again only discuss the most recent data. Each subsidiary was asked to rate its performance in comparison to other companies operating in the same industry. We looked at three different aspects of performance: market performance (profitability, sales growth, and market share), process performance (product quality, innovation, and productivity), and HRM performance (employee development and staff retention). Not surprisingly, most subsidiaries showed an acquiescence bias in this question. Very few subsidiaries claimed they performed worse than average. However, interesting differences are still apparent. Subsidiaries from the four different HQ countries differ in what they see as their best area of performance. For German and Japanese subsidiaries process performance rates highest, while for British and US subsidiaries this is HRM performance. Japanese and US subsidiaries rate their performance on market indicators as lowest, although in a cross-country comparison this performance is significantly higher for the

Variable

Performance 2002 data Market performance (1–7) Process performance (1–7) HRM performance (1–7) Significant differences Total significant differences 1995 Total significant differences 2002

Overview of Differences in Performance in 2002.

Mean Scores for Different Aspects of the HQ–Subsidiary Relationship in British, German, Japanese, and US MNCs

Difference between Difference between Difference between Difference British and German and British and US between Japanese German MNCs Japanese MNCs MNCs and US MNCs

British MNCs

German MNCs

Japanese MNCs

US MNCs

Sign. level

Sign. at .05/.10?

Sign. level

Sign. at .05/.10?

Sign. level

Sign. at .05/.10?

Sign. level

Sign. at .05/.10?

5.01

4.92

4.56

5.18

.761

No

.255

No

.446

No

.035

Yes

4.86

5.06

5.20

5.30

.366

No

.696

No

.010

Yes

.480

No

5.07

4.51

4.67

5.40

.044

Yes

.716

No

.101

No

.001

Yes

1/3

0/3

1/3

2/3

9/16 ¼ 56% 12/35 ¼ 34%

6/16 ¼ 38% 9/35 ¼ 26%

3/16 ¼ 19% 2/35 ¼ 6%

9/16 ¼ 56% 21/35 ¼ 60%

Headquarters–Subsidiary Relationships and Country-of-Origin Effect

Table 7.

33

34

ANNE-WIL HARZING AND NIELS NOORDERHAVEN

US than for Japan. German subsidiaries feel their lowest relative performance is in HRM, while for British subsidiaries this is in process performance. The largest differences between countries are in HRM performance with German and Japanese subsidiaries scoring significantly lower than British and US subsidiaries.

DISCUSSION AND CONCLUSION This study used primary data collected at some 30–50 subsidiaries from MNCs headquarted in four different countries and compared HQ– subsidiary relationships for two distinct time periods. By comparing MNCs from the same countries across two different time periods, we have been able to give a more comprehensive picture of possible changes in country-oforigin effects than earlier studies. Our results provide further support for the existence of unique country patterns, even for the most internationalized companies in the world. We therefore find a strong counter-argument against Ohmae’s (1990) suggestion of nationless corporations. Following Hu (1992) we think it would be better to describe MNCs as national firms with international operations instead. A clear conclusion of our study is that there are large differences in nearly all aspects of the HQ–subsidiary relationship between US and Japanese MNCs. These differences are persistent in that there are as many differences between MNCs from these two countries in 2002 as there were in 1995. British and US MNCs, however, are very similar in terms of their HQ– subsidiary relationships. In fact our 2002 survey shows even fewer differences than our 1995 survey. The lack of differences between US and British MNCs might seem surprising in the context of previous literature and recent studies conducted by Ferner et al. (see e.g. Ferner et al., 2004) which show that US MNCs are more centralized, standardized, and formalized in their HR practices than British MNCs. The tendency of US MNCs to favor transfer of their home country HR practices more so than Japanese and German MNCs was certainly confirmed in a recent study by Pudelko and Harzing (2007) that showed that subsidiaries of Japanese and German MNCs adopted US practices, while US subsidiaries showed a mix of localization and country-of-origin effect. However, we should note that all of these studies have focused only on HRM and industrial relations practices, while our current study takes a much broader perspective. As Pudelko and Harzing (2007) suggested MNCs might limit transfer of practices to their areas of core competence.

Headquarters–Subsidiary Relationships and Country-of-Origin Effect

35

A comparison of German MNCs with their British and Japanese counterparts provides a more mixed picture. Although German MNCs are different from both British and Japanese MNCs, they show a larger number of significant differences with their British than with their Japanese counterparts. This is almost as true in 2002 as it was in 1995. In addition, even where differences were not statistically significant, we discovered an overall pattern where Japanese and German MNCs clustered together, being quite distinct from their British and US counterparts. However, in terms of localization of subsidiary management (i.e. the reduction of expatriates in subsidiary top management) and the localization of production and R&D German MNCs do seem to have adapted to the Anglo-American practices of higher levels of localization. We could of course wonder to what extent the differences uncovered in this chapter are enduring and hence whether convergence might become more important in the 21st century. Lane (2000, 2001) suggested that German multinationals might have deviated from established societal patterns in the second part of the nineties. She mentioned that subsidiaries of German multinationals have been allocated more resources and granted more autonomy, with the organizations moving towards a decentralized network structure and subsidiaries becoming more embedded in the local environment through outsourcing and local recruitment of management. This certainly resonates with some of the findings in our 2002 study. However, we still find many differences between British and German MNCs and even more differences between US and Japanese MNCs. This finding suggests long-lasting stability of comparative differences over time. These differences clearly appear to be rooted in different postures and strategies, which have remained rather stable throughout an extensive period of industrialization and post-industrialization. They are the consequence of different conceptions of what the identity and comparative advantage of the firm should be built on: the product and engineering template in Germany and Japan, and differentiated marketing plus integrated financial management in Britain and the USA. Such postures and strategies lead to different paths of internationalization, and they are not necessarily specific for historically successive phases of internationalization. The British and US MNCs will therefore almost always seek to make the best profit in a conglomerate, which is more locally responsive, whereas the German and Japanese MNCs will strive towards a specialist technical template that can be implemented across locations. Consequently, we extend a strong plea for more empirical research into the country-of-origin effect for MNCs in general and the study of previously

36

ANNE-WIL HARZING AND NIELS NOORDERHAVEN

neglected MNCs of European origin in particular. A lack of empirical research in international business and management in general and a lack of non-US research in particular has created several myths (Harzing, 2002). Too often, MNCs have been presented as members of a species of their own. It is time to recognize that MNCs are, after all, firms, just like companies operating within the borders of a single state, and that they are susceptible to the same social influences as other firms. From some perspectives MNC can certainly be considered ‘‘harbingers of globalization’’ (Dicken, 1998), but from another perspective they are also products of the ‘‘administrative heritage’’ (Lubatkin et al., 1998) of their country of origin. Just like other companies, MNCs are after all products of the human mind.

REFERENCES Bartlett, C. A., & Ghoshal, S. (1989). Managing across borders: The transnational solution. Boston, MA: Harvard Business School Press. Child, J. (1984). Organization: A guide to problems and practice. London: Harper and Row. Dicken, P. (1998). Global shift: Transforming the world economy. London: Paul Chapman. Do¨rre, K. (1996). Globalstrategien von Unternehmen – ein Desintegrationspha¨nomen? Zu den Auswirkungen grenzu¨berschreitender Unternehmensaktivita¨ten auf die industriellen Beziehungen. SOFI Mitteilungen, 25, 43–70. Feely, A. J., & Harzing, A. W. K. (2003). Language management in multinational companies. Cross-Cultural Management: An International Journal, 10(2), 37–52. Ferner, A. (1997). Country of origin effects and human resource management in multinational companies. Human Resource Management Journal, 7, 19–37. Ferner, A., Almond, P., Clark, I., Colling, T., Edwards, T., Holden, L., & Muller-Camen, M. (2004). The Dynamics of Central Control and Subsidiary Autonomy: Case-Study Evidence from US MNCs in the UK. Organization Studies, 25(3), 363–391. Galbraith, J. R. (1973). Designing complex organizations. Reading, MA: Addison-Wesley. Ghoshal, S., & Nohria, N. (1989). Internal Differentiation within Multinational Corporations. Strategic Management Journal, 10(1989), 323–337. Giddens, A. (1999). Runaway World: How Globalization is Reshaping our Lives. London: Profile Books. Gupta, A. K., & Govindarajan, V. (2000). Knowledge flows within multinational corporations. Strategic Management Journal, 21(4), 473–496. Harzing, A. W. K. (1997). Response rates in international mail surveys: Results of a 22 country study. International Business Review, 6, 641–665. Harzing, A. W. K. (2001). Of bears, bumble-bees and spiders: The role of expatriates in controlling foreign subsidiaries. Journal of World Business, 36(4), 366–379. Harzing, A. W. K. (2002). Are our referencing errors undermining our scholarship and credibility? The case of expatriate failure rates. Journal of Organizational Behavior, 23(February), 127–148.

Headquarters–Subsidiary Relationships and Country-of-Origin Effect

37

Harzing, A. W. K., & Feely, A. J. (2008). The language barrier and its implications for HQ–subsidiary relationships. Cross-Cultural Management: An International Journal, 15(1), 49–60. Harzing, A. W. K., & Sorge, A. M. (2003). The relative impact of country-of-origin and universal contingencies on internationalization strategies and corporate control in multinational enterprises: World-wide and European perspectives. Organisation Studies, 24(2), 187–214. Hayden, A., & Edwards, T. (2001). The erosion of the country of origin effect; a case study of a Swedish multinational company. Relations Industrielles/Industrial Relations, 56, 116–140. Hennart, J.-F. (1991). Control in multinational firms: The role of price and hierarchy. Management International Review, 31(Special Issue), 71–96. Hirst, P., & Thompson, G. (1996). Globalisation in question. Cambridge: Polity Press. Hu, Y.-S. (1992). Global or stateless corporations are national firms with international operations. California Management Review (Winter), 107–126. Lane, C. (1998). European companies between globalization and localization: A comparison of internationalization strategies of British and German MNCs. Economy and Society, 27, 462–485. Lane, C. (2000). Globalisation and the German model of capitalism – erosion or survival? British Journal of Sociology, 51, 207–234. Lane, C. (2001). The emergence of German transnational companies and their impact on the domestic business system. In: G. Morgan, R. Whitley & P. H. Kristensen (Eds), The multinational firm: Organizing across institutional and national divides (pp. 69–96). Oxford: Oxford University Press. Lawrence, J. W., & Lorsch, P. R. (1967). Organization and environment. Harvard University Press. Lindholm, N. (1999–2000). National culture and performance management in MNC subsidiaries. International Studies of Management and Organization, 29, 45–66. Lubatkin, M., Calori, R., Very, P., & Veiga, J. F. (1998). Managing mergers across borders: A two-nation exploration of a nationally bound administrative heritage. Organization Science, 9, 670–684. March, J. G., & Simon, H. A. (1958). Organizations. New York: John Wiley and Sons, Inc. Martinez, J. I., & Jarillo, J. C. (1989). The evolution of research on coordination mechanisms in multinational corporations. Journal of International Business Studies, 20(3), 489–514. Mintzberg, H. (1983). Structure in fives: Designing effective organizations. Englewood Cliffs, NJ: Prentice-Hall. Morgan, G., & Kristensen, P.H. (2005). Constitutional orders in multinational firms: An institutionalist perspective on international management. Paper presented at the 65th Annual meeting of the Academy of Management, Honolulu, August 5–10. Ngo, H.-Y., Turban, D., Lau, C.-M., & Lui, S.-Y. (1998). Human resource practices and firm performance of multinational corporations: Influences of country origin. The International Journal of Human Resource Management, 9, 632–652. Nohria, N., & Ghoshal, S. (1994). Differentiated fit and shared values: Alternatives for managing headquarters–subsidiary relationship. Strategic Management Journal, 15, 491–502. Noorderhaven, N. G., & Harzing, A. W. (2003). The ‘‘country-of-origin effect’’ in multinational corporations: Sources, mechanisms and moderating conditions. Management International Review, 43(Special issue 2), 47–66. Ohmae, K. (1990). The borderless world: Power and strategy in the interlinked economy. London: Collins.

38

ANNE-WIL HARZING AND NIELS NOORDERHAVEN

Otterbeck, L. (1981). Concluding remarks and a review of subsidiary autonomy. In: L. Otterbeck (Ed.), The management of headquarters subsidiary relationships in multinational corporations (pp. 337–343). Aldershot, Hampshire: Gower. Ouchi, W. G. (1980). Markets, bureaucracies and clans, administrative science quarterly (Vol. 25, pp. 129–141). Pauly, L. W., & Reich, S. (1997). National structures and multinational corporate behavior: Enduring differences in the age of globalization. International Organization, 51, 1–30. Prowse, S. (1994). Corporate governance in an international perspective: A survey of corporate control mechanisms in the United States, the United Kingdom, Japan and Germany. BIS Economic Papers, No. 41 (July). Basle: Bank for International Settlements. Pudelko, M., & Harzing, A. W. K. (2007). Country-of-origin, localization or dominance effect? An empirical investigation of HRM practices in foreign subsidiaries. Human Resource Management, 46(4), 535–559. Ronen, S., & Shenkar, O. (1985). Clustering countries on attitudinal dimensions: A review and synthesis. Academy of Management Review, 10, 435–454. Ruigrok, W., & van Tulder, R. (1995). The logic of global restructuring. London: Routledge. So¨lvell, O., & Zander, I. (1995). Organization of the dynamic multinational enterprise; the home-based and the heterarchical MNE. International Studies of Management and Organization, 25(1–2), 17–38. Tregaskis, O. (1998). HRD in foreign MNEs. International Studies of Management and Organization, 28(1), 136–163. Whitley, R. (1998). Internationalization and varieties of capitalism: The limited effects of crossnational coordination of international activities on the nature of business systems. Review of International Political Economy, 5(3), 445–481. Whitley, R. (1999). How and why are international firms different? The consequences of crossborder managerial coordination for firm characteristics and behaviour. Paper presented at the 15th EGOS Colloquium, Warwick, July 4–6. World Bank. (2004). http://siteresources.worldbank.org/DATASTATISTICS/Resources/GDP.pdf

APPENDIX 1. MEASURES USED IN THE STUDY Control mechanisms can be defined as the instruments that are used to ensure that all units of the organization strive towards common organizational goals. Numerous control mechanisms have been identified. But following a synthesis of authors such as March and Simon (1958), Lawrence and Lorsch (1967), Child (1984), Galbraith (1973), Ouchi (1980), Mintzberg (1983), Bartlett and Ghoshal (1989), Martinez and Jarillo (1989), and Hennart (1991), the concept of control is mainly structured along two dimensions: directness and explicitness of control on the one axis, and impersonality of control on the other. This allows us to identify four major types of control mechanisms as summarized in Table A1. Based on the literature review, several constituent elements were defined for each of the four control mechanisms.

Headquarters–Subsidiary Relationships and Country-of-Origin Effect

Table A1.

Direct/explicit Indirect/implicit

39

Classification of control mechanisms on two dimensions. Personal/Cultural (Founded on Social Interaction)

Impersonal/Bureaucratic/ Technocratic (Founded on Instrumental Artifacts)

Centralization, direct supervision Socialization, informal communication, training and task forces

Standardization, formalization Output control, planning

To measure the various constituent elements of the different control mechanisms, we adapted and supplemented the questions that were used by Martinez and Jarillo (1989). We decided to combine the two impersonal control mechanisms as the data showed them to load on the same factor. Cronbach’s alpha reliabilities ranged from 0.68 to 0.74. In the 2002 study, we decided to differentiate the centralization aspect of direct personal control by asking respondents to assess – on a 5-point scale – the influence HQ would normally have on a range of issues varying from selection of suppliers to design of advertising for the local market. This measure was adapted from Otterbeck (1981). Exploratory factor analysis resulted in two clearly separated factors that reflected centralization of upstream activities (a ¼ 0.81) and centralization of downstream activities (a ¼ 0.67). Two questions were used to assess the presence of expatriates in a given subsidiary. These questions asked respectively for the nationality of the managing director and the number of top five (1995) or top ten (2002) jobs held by expatriates. The nationality of the managing director was recoded into 0 if the managing director was a local and 1 if the managing director was a parent country national. The small number of third country nationals was disregarded. The importance of the different functions of expatriation was probed with a series of single-item measures covering the major functions of expatriation (see also Harzing, 2001a–c). In the 2002 study, several of these functions (improving communication, knowledge transfer, training for future positions) were differentiated, looking at either interaction with HQ or with subsidiaries. Interdependence was operationalized using the percentage of intracompany sales. In the 1995 survey respondents were asked to differentiate between their purchases from or sales to HQ and subsidiaries, so that we could verify the relative importance of their interdependence with both HQ and other subsidiaries. Six answer categories were included: 0%, 1–25%, 26–50%, 51–75%, 76–99%, and 100%. The percentage of purchases from

40

ANNE-WIL HARZING AND NIELS NOORDERHAVEN

external suppliers was not measured directly, but calculated as the balance. In the 2002 study we refined this measures by asking respondents to estimate the percentage of their subsidiary’s input from different entities: HQ/ subsidiaries in the country of HQ, other subsidiaries in the same country or abroad, and external suppliers in the same country or abroad. The two measures for subsidiaries and suppliers were subsequently averaged. The 2002 study also included a second measure of interdependence focusing on knowledge flows. This measure was taken from Gupta and Govindarajan (2000). However, given the large number of constructs in our questionnaire, we decided to reduce their seven areas of knowledge flows to four: (1) product design, (2) marketing, (3) distribution, and (4) management systems and practices. Following Gupta and Govindarajan (2000), the respondent was asked to indicate on a scale from 1 to 7 the extent to which the subsidiary engaged in the transfer of knowledge and skills to and from HQ and other subsidiaries in each of the areas above. Cronbach’s a was 0.79 for both HQ flows and subsidiary flows. In the 1995 study, local responsiveness was measured with four items asking for the percentage of local R&D and local production incorporated in products sold by the subsidiary and the percentage of products and marketing that was substantially modified for the local markets. As for interdependence, six answer categories were created. In the 2002 study we refined this measures by asking respondents to estimate the percentage of local manufacturing and R&D directly, distributing 100% over five categories: this subsidiary, HQ/subsidiaries in the country of HQ, other subsidiaries in the same country or abroad, and external suppliers. Other aspects of local responsiveness (product/marketing modification and collaboration with local suppliers and customers) were measured with single-item measures on a 7-point Likert scale. Measures with regard to language barriers and policies were factual questions asking for the languages spoken at HQ and the subsidiary in question, the existence of a corporate language and the number of functions that were involved in cross-lingual communication. We also asked respondents to estimate the capability that subsidiary staff had in HQ and corporate language on a 7-point scale. With regard to performance we attempted to cover the broad spectrum of performance dimensions ranging from profitability to innovation and employee development. Eight items were included. Exploratory factor analysis separated these into three different aspects: market performance (profitability, sales growth, and market share, a ¼ 0.71), process performance (product quality, innovation, and productivity, a ¼ 0.67), and HRM performance (employee development and staff retention, a ¼ 0.75).

AN EXPLORATORY STUDY OF CULTURE DISTANCE AND PERCEPTIONS OF RELATIONAL RISK Susana Costa e Silva and Luciara Nardon ABSTRACT This chapter is part of a research project examining the role of culture and culture differences in foreign partnerships. We build on prior research on culture distance to explore the influence of perceptions of cultural differences on perceived relational risk. Perceived relational risk is defined here as the degree of satisfaction of being involved in business activities with nationals of a given country. Contrary to expectations, our analysis suggests that cultural differences are sometimes perceived as a desirable characteristic and may be associated with lower relational risk. We speculate that culture distance is an asymmetric construct in which the perception of a cultural difference may be interpreted as positive or negative depending on the perspective from which the reading is made and the nature of the task in which the perception is formed.

At the present time there is a greater need for effective international and cross-cultural communication, collaboration, and cooperation, not only for the effective practice of New Perspectives in International Business Research Progress in International Business Research, Volume 3, 41–58 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1745-8862/doi:10.1016/S1745-8862(08)03003-3

41

42

SUSANA COSTA E SILVA AND LUCIARA NARDON management but also for the betterment of the human condition. Ample evidence shows that cultures of the world are getting more and more interconnected and that the business world is becoming increasingly global. As economic borders come down, cultural barriers will most likely go up and present new challenges and opportunities in business. (House, 2004, p. 1)

INTRODUCTION As well described by Robert House (2004) in the opening quote, increasing globalization brings the need to cooperate and collaborate with people from various cultural backgrounds, increasing the range of opportunities available to managers and at the same time posing important challenges. Developing successful relationships with people from different cultures may be challenging for several reasons, including people’s tendency to have preconceived assumptions about how the world works, how individuals behave, and which behaviors are acceptable or unacceptable in a business setting. These ideas are largely influenced by our personal experiences and cultural background. We tend to approach intercultural situations using our own perceptions, beliefs, values, biases, and assumptions about what is likely to happen as a guide (Kluckhohn, 1954; Geertz, 1973; Hofstede, 1980, 1991; Trompenaars & Hampden-Turner, 1998; Schneider & Barsoux, 2003; Steers & Nardon, 2006). As a result, in international situations the consequences of our actions are frequently different than we expected or intended (Adler, 2002). The results can range from embarrassment to lost business opportunities. Managers are aware of these challenges and consciously or unconsciously take them into consideration before embarking in a business relationship with foreign counterparts. To this end, common sense suggests that small cultural differences are better than large ones, as higher degrees of difference are likely to result in more problems. This assumption has shaped most of the cross-cultural management literature, and is explicit in several studies using measures of culture distance to predict behavior. Following this line of thought one would expect the perceived relational risk of an interaction, that is the degree of satisfaction of being involved in business activities with nationals of a given country, would be higher in situations of smaller culture distance. However, contrary to expectations, our analysis suggests that cultural differences are sometimes perceived as a desirable characteristic and associated with lower perceived relational risk.

Exploratory Study of Culture Distance and Perceptions of Relational Risk

43

This chapter presents the results of an exploratory pilot study with Portuguese managers doing business internationally, exploring perceptions of cultural differences and relational risk. We will begin this chapter with a review of relevant literature on culture and culture distance. Then, we will present the results of our exploratory interviews and offer some insights on the role of culture distance in shaping relational risk assessment.

NATIONAL CULTURE AND CULTURE DISTANCE The word ‘‘culture’’ has been defined in several ways. Trompenaars (1993), for example, defines culture as the way in which a group of people solves problems and reconciles dilemmas. Clifford Geertz (1973) defines culture as the means by which people communicate, perpetuate, and develop their knowledge about attitudes towards life; culture is the fabric of meaning in terms of which people interpret their experience and guide their action. Kluckhohn’s (1954) approach defines culture as the collection of beliefs, values, behaviors, customs, and attitudes that distinguish the people of one society from another. Finally, Hofstede (1991) suggests that culture is the collective programming of the mind that distinguishes the members of one human group from another. Culture is the glue that ties a group or society together and signifies what they stand for. In both the personal and the business world, culture establishes the rules that govern how people and organizations operate. Cross-cultural management scholars have recognized that variance in these rules of behavior across nations are likely to have important implications for managers. To this end, significant research has focused on classifying cultures and identifying cultural dimensions that are meaningful to managers, in an attempt to map out meaningful differences. At present, there are several models available to examine the role of culture in organizations. These models focus on different aspects of societal beliefs, norms, or values. Cultural models have been proposed by Kluckhohn and Strodtbeck (1961), Hofstede (1980, 2001), Hall (1959, 1981), Trompenaars (1993), Schwartz (1994), and the GLOBE research team (House, Hanges, Javidan, Dorfman, & Gupta, 2004) and have offered useful templates for comparing management processes, human resource management (HRM) policies, and business strategies across national borders. Of these models, Hofstede’s (1980) has enjoyed the most popularity within the cross-cultural management field, as it was the first to focus on management and provide numerical indicators for various countries

44

SUSANA COSTA E SILVA AND LUCIARA NARDON

facilitating its use in empirical investigation of cultural influences on managerial behavior. For instance, Kirkman, Lowe, and Gibson’s (2006) review of 180 studies using Hofstede’s (1980, 2001) cultural values framework between 1980 and 2002 suggests that cultural values have been associated with change management, conflict management, decisionmaking, HRM, leadership, organization citizenship behavior, work-related attitudes, negotiation behavior, reward allocation, and individual behavior relating to group processes and personality, among others. Moreover, some authors focus on the cultural differences among nations and recognize that the degree of difference (measured through distance among variables) is likely to have an impact on business. This stream of thought focuses on identifying the role of cultural distance in several business issues (e.g. Kogut & Singh, 1988; Morosini, Shane, & Singh, 1998; Manev & Stevenson, 2001). Building on this literature, this research focuses on the way cultural differences are perceived and influence internationallevel business decisions. Further we review the culture distance literature.

CULTURAL DISTANCE The notion of cultural distance was first introduced by Beckermann (1956), and gained popularity within business studies within Scandinavian school, which observed that Swedish firms expanded to new markets progressively from lower to higher cultural distance (Johanson & Vahlne, 1977). In their seminal article, Kogut and Singh (1988) introduced a measure of cultural distance which can be used to empirically test the effect of cultural differences in organizational outcomes. The cultural distance index is constructed as an aggregation of Hofstede’s four original cultural dimensions (1980): uncertainty avoidance, individualism–collectivism, power distance, and masculinity–femininity. The cultural distance index is computed through the deviation along each of the four cultural dimensions between a country of focus and other countries as illustrated in Eq. (1). P4 hðI ij I iu Þ2i CDj ¼

i¼1

Vi

4

(1)

where, j is the country whose cultural distance is being measured; i, each of the four dimensions; and u the country of reference. This index was subsequently used in many international business studies investigating topics ranging from foreign direct investment,

Exploratory Study of Culture Distance and Perceptions of Relational Risk

45

headquarter–subsidiary relations, expatriate selection, and adjustment (see, for example, Benito & Gripsrud, 1992; Reuer & Tong, 2005). The notion of a cultural distance index is appealing as it eliminates the complexities usually associated with cultural issues, and provides a handy number to be used in empirical modeling. Following Kogut and Singh (1988), Morosini et al. (1998) proposed an adaptation on Kogut and Singh’s index to estimate the cultural distance between Italy and other countries based on Hofstede’s four dimensions of culture (Eq. (2)). Manev and Stevenson (2001) also used the work of Kogut and Singh (1988) as a base to define cultural distance as the degree to which cultural norms in one country differ from the ones of another. These authors based their research on managers with an expatriate status and reached the conclusion that a minor cultural distance allows the development of stronger instrumental ties between managers. Cultural distance was computed here as a Euclidian distance similar to Kogut and Singh’s index (Eq. (3)). vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi u 4 uX (2) CDj ¼ t ðI ij  I iI Þ2 i¼1

where, CDj is the cultural difference for the jth country; Iij, the Hofstede’s score for the ith cultural dimension and jth country; and I indicates Italy. vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi u 4 uX (3) CDij ¼ t ðDik  Djk Þ2 K¼1

where CDi is the cultural distance between managers i and j, and Dik and Djk are the indices for the kth dimension in is and js national cultures. Despite its popularity, the cultural distance index as originally developed by Kogut and Singh has received some important criticisms. For instance, in a review of the culture distance literature, Shenkar (2001) identified several conceptual and methodological concerns regarding this construct. He suggests that the cultural distance index has several hidden faulty assumptions built into it, namely: (1) an illusion of symmetry, or the notion that the role of the cultural distance is identical from the home and host countries point of view; (2) an illusion of stability, or the notion that culture is static and the distance remains constant over time; (3) an illusion of linearity, or the idea that the effect of cultural distance on business issues

46

SUSANA COSTA E SILVA AND LUCIARA NARDON

is linear; (4) an illusion of causality, or the assumption that cultural distance is the only determinant of distance that is relevant; (5) and illusion of discordance, or the notion that cultural differences are undesirable. He also argues that some methodological properties of the cultural distance index are problematic, such as: (1) the assumption of corporate homogeneity, which refers to the fact that the index does not account for organizational culture; (2) the assumption of spatial homogeneity, or the notion that countries are culturally uniform; (3) the assumption of equivalence, or the assumption that all four cultural variables included in the index are equally important. Finally, some argue that cultural distance is a cultural-level measure and therefore can only be used to investigate cultural-level phenomenon (Sousa & Bradley, 2006). Even though several authors use cultural distance and psychic distance interchangeably (e.g. Shoham & Albaum, 1995; Lee, 1998), recently Sousa and Bradley (2006) have argued that the two constructs are conceptually different. These authors argue that cultural distance is a national-level phenomenon while psychic distance is an individual level one. Psychic distance is the perception an individual has regarding the cultural difference between the home and foreign countries. As such, it is highly subjective and cannot be measured with factual indicators. Additionally, perceptions can be changed by increased experience with the foreign culture. However, they argue cultural distance refers to the cultural level of analysis and as such should only be applied to country-level studies. However, despite the above criticism, the Kogut and Singh index continues to be widely used, in part due to its convenience and in part due to the lack of a reliable alternative (Dahl, 2004). The Kogut and Singh Index is mainly used in studies focusing on culture’s influence on firm’s performance (see for instance, Barkema & Vermeulen, 1997; Lincoln, Hanada, & Olson, 1981; Pothukuchi, Damanpour, Choi, Chen, & Park, 2002), with contradictory findings. For instance, Larimo (2003) found that culture distance negatively affect performance, Salk and Brannen (2000) found no effect of culture distance on performance and Morosini and colleagues (1998) and Kessapidou and Varsakelis (2002) found that culture distance positively affects performance. The inconclusive findings of this line of study suggest that current conceptualizations of cultural distance do not fully represent the complexity of business across cultures. In this study we elaborate on the current literature on culture distance, focusing on the specific issue of relational risk.

Exploratory Study of Culture Distance and Perceptions of Relational Risk

47

CULTURE DISTANCE AND RELATIONAL RISK Scholars have long recognized that engaging in international partnerships brings new business opportunities, but also new challenges and risks. Whereas most of the research in this area emphasizes the advantages of partnering and highlights the need for mutual trust (see, for example, Kale, Singh, & Perlmutter, 2000; Moorman, Zaltman, & Deshpande´, 1992), the importance of trust is a result of the high risk associated with such international partnerships (Jeffries & Reed, 2000). In other words, international relationships carry risks, and these risks can be mitigated by trust. However, regardless of the level of trust in a relationship, risk is always present, as the attitude of trusting implies that there is risk. Thus, there is no trust if risk is not present. As noted by Nooteboom, Berger, and Noorderhaven (1997) ‘‘ . . . trust pays, but it also carries the risk of betrayal’’. Trust and risk are therefore important components when evaluating actors’ satisfaction with a partnership. Even though most research focuses on levels of trust in foreign partners, we argue that perceived relational risk is also an important indicator of satisfaction with a partnership. In other words, we argue that the level of satisfaction with a foreign partnership is dependent on the perception of relational risk. Following Das and Teng (1996), we propose that relational risk is an individual’s perception of the risks associated with cooperating with the alliance partner. This risk is subjective, and while it is influenced by objective characteristics of the situation, it is also influenced by each individual’s risk preferences, and past experiences (Nooteboom et al., 1997; Ring & Ven, 1992). We argue that the risk implicit in a relationship is reflected in the satisfaction retrieved from that relationship. In other words, we argue that the willingness to engage in a relationship with foreign nationals is influenced by the perception of risk involved in the relationship. Hence, the lower the risk perceived in a relationship, the higher the satisfaction with that partnership. Nevertheless, the problem of risk assessment is not independent from trust itself. In fact, we argue that risk besides being an important element in evaluating satisfaction along with trust, is also an important component of trust. Scholars have long recognized that interpersonal trust is important for business, because it makes actors more willing to collaborate, increases the likelihood of success (see for instance Madhok, 1995; Ring & Ven, 1992; Smith, Carroll, & Ashford, 1995), and the level of satisfaction of partners (Gulati, 1998; Powell, 1996; Sako, 1998; Zaheer, McEvily & Perrone, 1998b). Nowadays, as a result of increasing globalization, ease of

48

SUSANA COSTA E SILVA AND LUCIARA NARDON

communication and transportation, parties from different nation states, cultures and languages are increasingly engaging in alliances and other business relationships (Ring & Ven, 1992). These relationships require ‘‘a much better understanding of how trust is developed, or manifested, in different cultures ( . . . )’’ (Ring & Ven, 1992, p. 496). It will also require an understanding on how trust is developed and manifested across cultures. The trust one has in a foreign counterpart is influenced by opposing forces. On one hand, the propensity to trust and attitudes toward cooperation are likely to vary across cultures (Steensma, Marino, & Weaver, 2000; Shane, 1992; Hill, 1990; Fukuyama, 1996), suggesting that people from some cultures may be more likely to trust foreigners than others. However, trust is facilitated by shared values and beliefs, so one would assume that it is easier to trust people from similar cultures than from distant ones. The trust literature suggests that trust is built upon different pillars: actor willingness to trust (Aulakh, Kotabe, & Sahay, 1996), belief on integrity and fairness (Zaheer, McEvily, & Perrone, 1998a), reliability, or likelihood that partner will honor commitments (Anderson & Weitz, 1989), and risk and vulnerability (Bromiley & Cummings, 1995; Lewis & Weigert, 1985; Mayer, Davis, & Schoorman, 1995). As discussed above, risk and vulnerability are key components of trust, as trust only exists if partners are aware to the possibility of being failed. In other words, when partner in a relationship are vulnerable regarding the other part trust is important (Sabel, 1993; Mayer et al., 1995). In an increasingly global economy, establishing international alliances became imperative for some firms survival and strategic for their continuation. In such environment differences in values, beliefs, assumptions, and contextual realities may at times hinder and at times facilitate developing trust among international partners. In this research we focus on the risk element of trust by exploring the perception of relational risk with foreign counterparts by Portuguese managers.

RESEARCH METHODS AND ANALYSIS This is the first step of a larger research project exploring perceptions of cultural differences and trust. At this stage, we have conducted five exploratory pilot interviews with Portuguese managers engaged in alliances with foreign partners, in order to understand their preferences in terms of the cultural background of business partners. These interviews were

Exploratory Study of Culture Distance and Perceptions of Relational Risk

49

semi-structured to allow personal, attitudinal, and value-laden material to emerge (Jankowicz, 1991). In this study we focus on how individuals from one culture (specifically, Portuguese) perceive the risk of engaging in business relationships with individuals from other cultures. With this analysis we want to uncover how cultural differences are perceived and the type of impact they have on relational risk assessment. We do not address in this study the role of culture in influencing trusting behavior. Participants were asked to name, based on their experience, the nationalities (cultures) they most and least liked to do business with. We assume that such preferences stem from the perceived relational risk, that is the ‘‘liking to do business’’ reflects a perceived low relational risk. To explore the effects of the transaction itself on their preference, we also asked participants to name the nationalities with which they liked most to engage in social relationships. Find later a brief summary of the interview data as it refers to business relationships, and the respective cultural distance index, as defined by Kogut and Singh. The pilot interviews reveal an interesting and surprising pattern as we may see in Fig. 1. Contrary to expectations, the five Portuguese managers interviewed prefer to do business with foreign partners from distant cultures, which we interpret as a perception of lower relational risk. Likewise, four of the interviewees dislike most to do business with partners of closer cultures. Interestingly, when enquired about their preferences in doing business with the nationality they like the most and their own people, they all preferred the foreign partner. These findings suggest that a measure of distance alone is not a good explanation in any direction (Tables 1 and 2). However, cultural differences were recognized as important criteria in assessing their partners, but they were not always considered negative. Instead, the perception of the other’s culture was used as an evaluative component comparing to one’s own culture. Some cultural characteristics different from one’s own were considered desirable. For example, the Portuguese culture is usually described as polychronic with regards to their perception of time (Hall, 1959, 1981; Steers and Nardon, 2006), meaning that punctuality is not considered a priority as time is seen as flexible, fluid, and relative. Yet, when asked what they disliked about their foreign counterparts they said ‘‘they are frequently late’’ and ‘‘they often miss deadlines’’. Similarly, Portuguese people tend to prefer indirect communication styles, focusing on the context of the communication to interpret messages, what Hall (1959, 1981) calls high-context communication. Yet, when asked why they did not like to work with Brazilians, Spaniards, and

50

SUSANA COSTA E SILVA AND LUCIARA NARDON Preferences in terms of socialization

Spain

Brazil

PT speaking African C.

Italy

Switzer land

Japan

USA

Sweden

China

Countries ranked according to their national cultural distance (based on Kogut & Singh Index) towards Portugal

Preferences in terms of doing business Note: The X axis tries to give an idea of the national cultural distance between Portugal and the referred countries. Symbols (faces and glasses) represent number of responses in our pilot interviews.

Nationalities with whom PT like to do business

Nationalities with whom PT like to socialize

Nationalities with whom PT like to socialize and don't like to do business

Nationalities that get no consensus regarding PT preferences of doing business

Fig. 1. Preferred Countries to do Business and to Socialize in Terms of Perceived Risk. Note: The X-axis Tries to Give an Idea of the National Cultural Distance between Portugal and the Referred Countries. Symbols (Faces and Glasses) Represent Number of Responses in Our Pilot Interviews.

Italians, the managers in our pilot interviews made comments such as ‘‘they don’t always follow through on what they say’’, ‘‘they speak indirectly’’, and ‘‘they do not say what they think’’. Noticing these contradictions one manager said ‘‘they are very much like us’’. Whether they openly

51

Exploratory Study of Culture Distance and Perceptions of Relational Risk

Table 1. Manager

1 2 3 4 5

Revealed Preferences to do Business.

Countries in Which Does Business

Sweden, China, Ireland, Spain, Japan USA, Germany, Italy, Spain Japan, China, Brazil, Russia China, Spain, Portuguese speaking African countries France, Germany, Switzerland, Spain, Italy

Preferred Nations

Sweden

Kogut and Least Kogut and Singh’s Liked Singh’s Index Nations Index 4.17

USA Japan Chinaa

China

4.00 Italy 3.29 Brazil NA (high) Spain

Switzerland

3.12

Spain

NA (high) 2.50 0.64 0.47 0.47

a

Even though we do not have equivalent information to calculate Kogut and Singh’s index for China, culture studies on China and Portugal suggest a high degree of national cultural difference.

Table 2. Manager

1 2 3 4 5

Revealed Preferences to Socialize.

Countries in Which Does Business

Preferred Nations

Kogut and Singh’s Index

Sweden, China, Ireland, Spain, Japan USA, Germany, Italy, Spain Japan, China, Brazil, Russia China, Spain, Portuguese speaking African countries France, Germany, Switzerland, Spain, Italy

Spain Spain Brazil African Portuguese speaking Countries Italy

0.47 0.47 0.64 1.35 2.5

acknowledged or not, these managers seemed to think that people from cultures ‘‘like ours’’ were less trustworthy and therefore the risk involved in these partnerships was considered to be higher. This is made more evident when these managers stated they prefer to do business with some foreign cultures rather than with their own people. However, when asked about the characteristics of the partners they appreciated, managers’ comments highlighted cultural characteristics that were different from their own, such as their direct communication style, assertiveness, and tendency to follow through on their words. Here again managers mentioned ‘‘they are different from us’’ as a positive characteristic. However, just being different is not enough. For instance, one manager disliked working with the Chinese because they were considered too intricate

52

SUSANA COSTA E SILVA AND LUCIARA NARDON

and difficult to understand. In this case, cultural differences were perceived as negative and difficult to cope with. Yet, another manager liked to do business with the Chinese because he reached a stage of knowledge about them that facilitates business. He sates ‘‘it took me more than two years to get to know how to negotiate with them, but now that I know how they are, negotiation is easy’’, which suggests that learning about the other culture may bridge possible gaps between culturally distant partners. While this data is far from conclusive, it does suggest a learning effect on dealing with culture. It is possible that the impact of cultural differences decrease with exposure to the culture, and that the willingness to trust the other is related to understanding how the other operates. This may explain why low-context, monochronic cultures were appreciated by this sample. Even though these cultural characteristics are different from the Portuguese, they are perceived as more predictable, and are easier for an outsider to understand. Take for instance low-context communication (Hall, 1959, 1981). Lowcontext communicators rely on words to deliver the message, while highcontext communicators rely on other cues, such as the parties involved, body language, and the context itself. High-context communicators frequently speak indirectly, and ‘‘do not say what they mean’’ as one of our informants suggested. For someone not immersed in the context it is difficult to comprehend the whole meaning of the communication as contextual cues are easily missed. However, communicating with lowcontext communicators is much easier as ‘‘what they say is what they mean’’. The same can be said about time orientation (Hall, 1959, 1981). While individuals in polychronic cultures expect people to be late, there is a tacit understanding of when things will be done. When in a foreign polychronic environment this understanding is not there, and there is a sense of chaos and confusion by not knowing what will happen when, as time perceptions are not clearly spelled out. However, it is easy for a polychronic culture to understand a monochronic person doing things ‘‘ontime’’. While this data is still exploratory, it suggests that some cultural characteristics may facilitate ‘‘being trusted’’, or offer less risk, suggesting that culture not only influences how willing one is to trust the other, but also how trustworthy one is thought to be. Therefore, our preliminary proposition: Proposition 1. Some cultural characteristics are perceived as less risky than others.

Exploratory Study of Culture Distance and Perceptions of Relational Risk

53

Proposition 1a. Low-context cultures are perceived as less risky than highcontext cultures. Proposition 1b. Monochronic cultures are perceived as less risky than polychronic cultures. Interestingly, however, when the same managers were asked about their preferences for social purposes, their preferences changed to cultures with small cultural distance, such as Italy, Spain, Brazil, and Portuguese speaking African countries. Here, managers claimed they feel more ‘‘at home’’ with such cultures and find their company more enjoyable. We speculate that the variance in preference is related to the task at hand. Cultural distance in itself is not necessarily an obstacle, but may be perceived sometimes as a positive, sometimes as negative, depending on the nature of the task and the purpose of the relationship. This is in line with Shenkar’s (2001) criticism that current cultural distance measures erroneously assume that the lack of cultural fit results in an obstacle to the transaction. However, he argues, different aspects of culture may be more or less critical to operations, and some cultural differences may be complementary, while others may be irrelevant. Notice, for instance that the managers in our pilot study did not mention variations in power distribution or different levels of collectivism and individualism. Presumably, these cultural aspects were not relevant in their dealings. Instead, communication styles, time orientation, and to some degree rule orientation were considered more critical by this sample when doing business. However, when the subject was socialization, cultural similarity was considered more important. Therefore, we suggest: Proposition 2. The perception of foreign partner’s relational risk is task dependent.

DISCUSSION Our pilot interviews suggest that Portuguese managers’ assessment of foreign partners’ relational risk is influenced by perceptions of national cultural differences, but that the amount of national cultural difference is not the most important criteria. Rather, the type and direction of these differences coupled with the nature of the task or purpose of interaction are more important. In other words, cultural differences are sometimes

54

SUSANA COSTA E SILVA AND LUCIARA NARDON

perceived as desirable, and sometimes perceived as undesirable, depending on the task characteristics. Based on limited data, we speculate that some cultural characteristics are perceived as more desirable than others, and those may be more important than any measure of difference. Even though our current sample is limited to Portuguese managers, we speculate that in some situations one party may find the other trustworthy but not be reciprocated. That is, cultural differences are not symmetric (Shenkar, 2001) in the sense that the same cultural variable may have different roles to the cultures involved. Fig. 2 illustrates the relationship between cultural differences and relational risk. We speculate that a foreign nation’s cultural characteristics combined with the nature of the task are the main drivers of perceived relational risk, but this relationship is moderated by the assessor’s home culture. In other words, an individual’s assessment of a partner’s relational risk, will be determined by a combination of the partner’s cultural characteristics and the task at hand – or the purpose of the relationship. However, we speculate that one’s own culture of reference is likely to play a role in moderating this relationship (for instance, a culturally based tendency to trust more or less, to accept different perceptions, or to value particular characteristics). Moreover, the role of one’s own culture is not in defining a degree of ‘‘difference’’ or cultural distance, but in providing a different perspective and expectations regarding which characteristics are desirable. For example, if the task at hand asks for punctuality and precise time management, Portuguese managers may perceive the relationship with Swedish managers as a low risk one, as Swedish people are usually associated with a culture of monochronic behavior. Thus, based on how Portuguese valuate Swedish national features and on the nature of the task,

Task Perceived Relational Risk Cultural Characteristics Foreign Country Cultural Characteristics Home Country

Fig. 2.

Cultural Differences and Perception of Relational Risk.

Exploratory Study of Culture Distance and Perceptions of Relational Risk

55

a certain relational risk is asserted to this relationship. Nevertheless, the meaning of ‘‘being on-time’’ is in itself influenced by the cultural characteristics of the home country, and moderates the influence of cultural characteristics on relational risk assessment.

CONCLUSION In this chapter we presented the results of an exploratory pilot study with Portuguese managers doing business internationally, exploring perceptions of cultural differences and relational risk. Contrary to expectations, we found that cultural differences were sometimes perceived positively and associated with lower relational risk. Based on this finding, we suggest that culture distance is an asymmetric construct in which the perception of a cultural difference may be interpreted as positive or negative depending on the perspective from which the reading is made and the nature of the task in which the perception is formed. Despite the exploratory nature of this research, the findings presented here have important implications. Within the international business field, several issues are thought to be influenced by cultural differences. Researchers attempting to identify and measure the influence of cultural differences on business phenomenon need to be cognizant of the variations in interpretation of these differences. Cultural differences may be perceived differently by different sides, and be considered positive or negative depending on the task at hand. A blind application of cultural distance indexes to the investigation of managerial behavior may lead to contradictory results.

REFERENCES Adler, N. J. (2002). International dimensions of organizational behavior (4th ed.). Cincinnati, OH: Southwestern. Anderson, E., & Weitz, B. (1989). Determinants of continuity in conventional industrial channel dyads. Marketing Science, 4(Fall), 310–323. Aulakh, P. S., Kotabe, M., & Sahay, A. (1996). Trust and performance in cross-border marketing partnerships: A behavioral approach. Journal of International Business Studies, 27(5), 1005–1032. Barkema, H. G., & Vermeulen, F. (1997). What differences in the cultural background of partners are detrimental for international joint ventures? Journal of International Business Research, 28(4), 845–864.

56

SUSANA COSTA E SILVA AND LUCIARA NARDON

Beckermann, W. (1956). Distance and the pattern of intra-European trade. Review of Economics and Statistics, 28, 31–40. Benito, G., & Gripsrud, G. (1992). The expansion of foreign direct investments: Discrete rational location choices or a cultural learning process? Journal of International Business Studies, 23(3), 461–476. Bromiley, P., & Cummings, L. L. (1995). Transaction costs in organizations with trust. In: R. Bies, B. Sheppard & R. Lewicki (Eds), Research on negotiation in organizations. Greenwich, CT: JAI Press. Dahl, S. (2004). Intercultural research: The current state of knowledge (Vol. 21). London: Middlesex University – Middlesex University Business School. Das, T. K., & Teng, B.-S. (1996). Risk types and inter-firm alliance structures. Journal of Management Studies, 33(6), 827–843. Fukuyama, F. (1996). Confianc- a – valores sociais e criac- a˜o de prosperidade (C. L. Borges, Trans.). Lisboa: Gradiva. Geertz, C. (1973). The interpretation of cultures. New York: Basic Books. Gulati, R. (1998). Alliances and networks. Strategic Management Journal, 19(SI), 293–317. Hall, E. T. (1959, 1981). The silent language. New York: Doubleday. Hill, C. W. L. (1990). Cooperation, opportunism and the invisible hand: Implications for transaction cost theory. Academy of Management Review, 15(3), 500–513. Hofstede, G. (1980). Culture’s consequences: International differences in work-related values. Beverly Hills, CA: Sage. Hofstede, G. (1991). Cultures and organizations: Software of the mind. London: McGrawHill. House, R. (2004). Introduction. In: R. J. House, P. J. Hanges, M. Javidan, P. W. Dorfman & V. Gupta (Eds), Culture, leadership and organizations: The GLOBE study of 62 societies. Thousand Oaks, CA: Sage Publications. House, R. J., Hanges, P. J., Javidan, M., Dorfman, P. W., & Gupta, V. (2004). Culture, leadership and organizations: The GLOBE study of 62 societies. Thousand Oaks, CA: Sage. Jankowicz, A. D. (1991). Business research projects for students. London: Chapman and Hall. Jeffries, F. L., & Reed, R. (2000). Trust and adaptation in relational contracting. The Academy of Journal Review, 25(4), 873–882. Johanson, J., & Vahlne, J.-E. (1977). The internationalization process of the firm - a model of knowledge development and increasing foreign market commitments. Journal of International Business Studies, 8(1), 23–32. Kale, P., Singh, H., & Perlmutter, H. (2000). Learning and protection of proprietary assets in strategic alliances: Building relational capital. Strategic Management Journal, 21(3), 217–237. Kessapidou, S., & Varsakelis, N. C. (2002). The impact of national culture on international business performance: The case of foreign firms in Greece. European Business Review, 14(4), 268–275. Kirkman, B. L., Lowe, K. B., & Gibson, C. B. (2006). A quarter century of culture’s consequences: A review of empirical research incorporating Hofstede’s cultural values framework. Journal of International Business Studies, 37(3), 285–320. Kluckhohn, F. (1954). Culture and behavior. In: G. Lindzey (Ed.), Handbook of social psychology (Vol. 2, pp. 921–976). Cambridge, MA: Addison-Wesley. Kluckhohn, F., & Strodtbeck, F. (1961). Variations in value orientations. Evanston, Illinois: Peterson Row.

Exploratory Study of Culture Distance and Perceptions of Relational Risk

57

Kogut, B., & Singh, H. (1988). The effect of national culture on choice of entry mode. Journal of International Business Studies, 19(3), 411–432. Larimo, J. (2003). International joint ventures strategies and performance in Asian countries. Paper presented at the Strategies for sustainable globalization, Bangkok. Lee, D.-J. (1998). The effect of cultural distance on the relational exchange between exporters and importers: The case of Australian exporters. Journal of Global Marketing, 11(4), 7–22. Lewis, J. D., & Weigert, A. (1985). Trust as a social reality. Social Forces, 63(4), 967–985. Lincoln, J. R., Hanada, M., & Olson, J. (1981). Cultural orientations and individual reactions to organizations: A study of employees of Japanese-owned firms. Administrative Science Quarterly, 26(1), 93–115. Madhok, A. (1995). Revisiting multinational firm’s tolerance for joint ventures: A trust-based approach. Journal of International Business Studies, 26(1), 117–137. Manev, I. M., & Stevenson, W. B. (2001). Nationality, cultural distance, and expatriate status: Effects on the managerial network in a multinational enterprise. Journal of International Business Studies, 32(2), 285–303. Mayer, R., Davis, J., & Schoorman, D. (1995). An integration model of organizational trust. Academy of Management Review, 20(3), 709–734. Moorman, C., Zaltman, G., & Deshpande´, R. (1992). Relationships between providers and users of market research: The dynamics of trust within and between organizations. Journal of Marketing Research, 29(3), 314–329. Morosini, P., Shane, S., & Singh, H. (1998). National cultural distance and cross-border acquisition performance. Journal of International Business Studies, 29(1), 137–158. Nooteboom, B., Berger, H., & Noorderhaven, N. (1997). Effects of trust and governance on relational risk. Academy of Management Journal, 40(2), 308–338. Pothukuchi, V., Damanpour, F., Choi, J., Chen, C. C., & Park, S. H. (2002). National and organizational culture differences and international joint venture performance. Journal of International Business Studies, 33(2), 243–265. Powell, W. W. (1996). Trust-based forms of governance. In: R. M. Kramer & T. R. Tyler (Eds), Trust in organizations (pp. 51–67). Newbury Park, CA: Sage. Reuer, J. J., & Tong, T. W. (2005). Real options in international joint ventures. Journal of Management, 31(3), 403–423. Ring, P. S., & Ven, A. H. V. D. (1992). Structuring cooperative relationships between organizations. Strategic Management Journal, 13(7), 483–498. Sabel, C. F. (1993). Studied trust: Building new forms of cooperation in a volatile economy. Human Relations, 46(9), 1133–1170. Sako, M. (1998). Does trust improve business performance? In: C. Lane & R. Bachmann (Eds), Trust within and between organizations (pp. 88–117). Oxford: Oxford University Press. Salk, J. E., & Brannen, M. Y. (2000). National culture, networks, and individual influence in a multinational management team. Academy of Management Journal, 43(2), 191–202. Schneider, S., & Barsoux, J. (2003). Managing across cultures. London: Prentice Hall. Schwartz, S. (1994). Beyond individualism-collectivism: New cultural dimensions of values. In: H. C. T. U. Kim, C. Kagitcibasi, S. C. Choi & G. Yoon (Eds), Individualism and collectivism: Theory, method and application (pp. 77–119). Thousand Oaks, CA: Sage. Shane, S. (1992). Why do some societies invent more than others? Journal of Business Venturing, 7(1), 29–46.

58

SUSANA COSTA E SILVA AND LUCIARA NARDON

Shenkar, O. (2001). Cultural distance revisited: Towards a more rigorous conceptualization and measurement of cultural differences. Journal of International Business Studies, 32(3), 519–535. Shoham, A., & Albaum, G. S. (1995). Reducing the impact of barriers to exporting: A managerial perspective. Journal of International Marketing, 3(4), 85–105. Smith, K. G., Carroll, S. J., & Ashford, S. (1995). Intra- and interorganizational cooperation: Toward a research agenda. Academy of Management Journal, 38(1), 7–22. Sousa, C. M. P., & Bradley, F. (2006). Cultural distance and psychic distance: Two peas in a pod? Journal of International Marketing, 14(1), 49–70. Steensma, H. K., Marino, L., & Weaver, K. M. (2000). Attitudes towards cooperative strategies: A cross-cultural analysis of entrepreneurs. Journal of International Business Studies, 31(4), 591–609. Steers, R. M., & Nardon, L. (2006). Managing in the Global Economy. Armonk, New York: M.E. Sharpe. Trompenaars, F. (1993). Riding the waves of culture: Understanding cultural diversity in business. London: Economist Books. Trompenaars, F., & Hampden-Turner, C. (1998). Riding the waves of culture: Understanding diversity in global business. New York: McGraw Hill. Zaheer, A., McEvily, B., & Perrone, V. (1998a). Does trust matter? Exploring the effects of interorganizational and interpersonal trust on performance. Organization Science, 9(2), 141–159. Zaheer, A., McEvily, B., & Perrone, V. (1998b). The strategic value of buyer–supplier relationships. International Journal of Purchasing and Materials, 34(3), 20–26.

DRIVERS OF INTERPERSONAL AND INTER-UNIT TRUST IN MULTINATIONAL CORPORATIONS Kristiina Ma¨kela¨, Wilhelm Barner-Rasmussen and Ingmar Bjo¨rkman ABSTRACT Purpose – This chapter explores the determinants of trust, a manifestation of the relational dimension of social capital, between interaction partners in multinational corporations at interpersonal and inter-unit levels of analysis. Methodology – The study is based on two quantitative data sets from the Finnish subsidiaries of foreign MNCs, one at the individual level and another at the unit level (n ¼ 265/102). Findings – Our results indicate that the drivers of trust exhibit similar patterns across both levels of analysis, but are stronger at the interpersonal level. Trust was significantly and positively related to the length of the relationship between the two individuals or units, and to the frequency of the communication between them whereas it was found to be unrelated with cultural distance. Limitations – The key limitations of the study were as follows. First, the samples at both levels were relatively small, a factor that may partly explain why some of the statistical relationships were relatively weak. New Perspectives in International Business Research Progress in International Business Research, Volume 3, 59–80 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1745-8862/doi:10.1016/S1745-8862(08)03004-5

59

60

KRISTIINA MA¨KELA¨ ET AL.

Second, the study was carried out in one location only, and our findings need to be corroborated in other cultural settings. Third, we only examined the level of trust from one side of the relationship, as a dyadic analysis was not possible with the present data. Practical implications – For practicing managers, the main message from our research is that communication frequency and the length of the relationship matter for the relational social capital that exists within MNCs – both at the interpersonal and unit levels.

1. INTRODUCTION A key contribution of social capital theory (Bourdieu, 1983; Coleman, 1988; Burt, 1992; Putnam, 1995) is the observation that a social actor’s actual or latent connections to other actors constitute a type of potentially valuable ‘capital’, implying that networks which have emerged for one purpose can become assets which can also be used for other purposes. Consequently, a number of scholars have adopted social capital theory as an intellectual framework for examining different aspects of corporate life (e.g., Adler & Kwon, 2002; Bolino, Turnley, & Bloodgood, 2002; Inkpen & Tsang, 2005; Leana & Van Buren, 1999; McFayden & Cannella, 2004; Nahapiet & Ghoshal, 1998; Oh, Labianca, & Chung, 2006; Tsai & Ghoshal, 1998). For example, Nahapiet and Ghoshal (1998) and Tsai and Ghoshal (1998) have examined how the structural, relational and cognitive dimensions of social capital influence the creation of intellectual capital and consequently competitive advantage of multinational corporations (MNCs). While the growing use of social-capital-based frameworks has broadened the understanding of intra-MNC relationships, we still have a less than complete picture of the drivers and dynamics of particularly the relational dimension of social capital. First, while MNCs have provided the context for key empirical work on social capital (e.g., Kostova & Roth, 2002; Tsai & Ghoshal, 1998), there is only limited large-scale research on factors that are associated with social capital within the MNC. Second, previous research has typically examined social capital either between MNC units or alternatively between individual managers and empirical evidence on how these two levels of analysis relate to each other is extremely limited. Theoretical contributions include studies by Bolino et al. (2002) and Kostova and Roth (2003), but we are not aware of any empirical work explicitly addressing this issue. As it has been forcefully argued that inter-unit relationships are a function of

Drivers of Interpersonal and Inter-Unit Trust in Multinational Corporations

61

interpersonal relationships (Brass, Galaskiewicz, Greve, & Tsai, 2004; Kostova & Roth, 2003), and the centrality of a given MNC unit may vary as a result of the connections individual managers hold to other units (Brass et al., 2004), this constitutes a significant gap in the literature. The lack of research concerning factors associated with relational social capital across different levels of analysis has also been noted in recent reviews (Schoorman, Mayer, & Davis, 2007). In this study, we focus on trust as a key manifestation of the relational dimension of social capital. More specifically, we examine factors associated with trust both (i) between individual managers working in the different units of the same MNC and (ii) between two units belonging to the same MNC. The chapter is structured as follows. First, we briefly review the concept of social capital in general and its relational dimension (trust) in particular as treated in the management literature, and develop hypotheses on factors that may facilitate or hinder the emergence of relational social capital between units and individuals within MNCs. We then describe our two samples and the statistical methods used to test our hypotheses. Finally, we present our results and discuss their relevance for both theory and practice.

2. FRAME OF REFERENCE 2.1. Social Capital and Its Dimensions The concept of social capital (Bourdieu, 1983; Burt, 1992; Coleman, 1988; Putnam, 1995), in broad terms referring to assets embedded in relationships, has recently attracted considerable attention in the social sciences. Within the bounds of this general idea, social capital has been defined and conceptualized in a number of ways (for a review and discussion, see e.g., Adler & Kwon, 2002). In this chapter, we adopt Nahapiet and Ghoshal’s (1998, p. 243; based on Bourdieu, 1983; and Putnam, 1995) definition of ‘the sum of the actual and potential resources embedded within, available through, and derived from the network of relationships possessed by an individual or social unit’. Within the field of management, the conceptualization by Nahapiet and Ghoshal (1998) has been argued to be particularly useful as it integrates many previously divergent strands of the literature (Adler & Kwon, 2002; Bolino et al., 2002; Inkpen & Tsang, 2005). They view social capital as having three overlapping and interlinked, yet distinct dimensions. The structural dimension is mainly concerned with physical linkages between

62

KRISTIINA MA¨KELA¨ ET AL.

people or units, such as network ties between actors; the pattern of ties in terms of their density, connectivity, or hierarchy; and the existence of networks created for one purpose that may be used for another (these have been the key focus of social network research, see e.g., Kildruff & Tsai, 2003; Lin, 2001). The relational dimension focuses on personal relationships and relations of mutual respect that individuals have developed through a history of interactions. It includes aspects such as trust and trustworthiness, norms and sanctions, obligations and expectations, and identity and identification (Nahapiet & Ghoshal, 1998; Tsai & Ghoshal, 1998). Finally, the cognitive dimension encompasses organizational phenomena such as shared representations, interpretations, language, codes, narratives, and systems of meaning among parties (Nahapiet & Ghoshal, 1998). A key contribution of the Nahapiet and Ghoshal (1998) approach in the context of international management research is to highlight the importance of the relational and cognitive dimensions. In earlier research, relational and cognitive linkages were largely inferred from the more easily quantifiable structural linkages, which have repeatedly been examined under headings such as ‘inter-unit interaction’ or ‘mechanisms of control and coordination’ (see, e.g., Ghoshal, Korine, & Szulanski, 1994; Gupta & Govindarajan, 2000). In this chapter, we focus on trust as a key aspect of the relational dimension of social capital, following Tsai and Ghoshal (1998).1 While the results of previous studies are not completely unambiguous (see BarnerRasmussen & Bjo¨rkman, 2007), empirical research (e.g., Tsai & Ghoshal, 1998) suggests that the relational dimension of social capital may be particularly relevant for knowledge exchange and combination, which in turn has been forcefully argued to play a key role for the competitiveness of the modern-day MNC (Grant, 1996; Kogut & Zander, 1992; Doz, Santos, & Williamson, 2001). Consequently, MNCs have reported to be under strong pressure to learn how to systematically manage and promote organizational trust (Kostova & Roth, 2003; Inkpen & Tsang, 2005). Tsai and Ghoshal (1998) were the first to apply the Nahapiet and Ghoshal (1998) framework to the empirical context of multinational corporations. They examined relationships between the three dimensions and the patterns of resource exchange and combination inside 15 units of a large MNC with results indicating that intra-MNC resource exchange and combination indeed increased as the levels of social capital rose. Of particular relevance for the present study is that they found a significant positive relationship between the relational dimension of social capital and resource exchange and combination. Other studies have also found evidence that high levels of trust or perceived trustworthiness facilitate collaboration, coordination of tasks

Drivers of Interpersonal and Inter-Unit Trust in Multinational Corporations

63

and knowledge flows in various contexts (Jones & George, 1998; McEvily, Perrone, & Zaheer, 2003; Uzzi, 1997; Uzzi & Lancaster, 2003; Zaheer, McEvily, & Perrone, 1998). Based on this discussion, we can now proceed to develop hypotheses concerning some of the drivers of trust in the MNC context, considering both inter-unit and interpersonal levels of analysis.

2.2. Trust and Relationship Length Because trust is a central component of the relational dimension of social capital, research on relational social capital naturally interfaces with the rich body of research on trust in organizational settings. In this literature there is a general agreement that relational social capital tends to develop between two parties over time through social interactions (Gulati, 1995; Tsai & Ghoshal, 1998). Sources of trust include, for example, ‘familiarity, shared experience y fulfilled promises, and demonstrations of non-exploitation of vulnerability’ (Meyerson, Weick, & Kramer, 1996, p. 167), all of which take relatively long to grow. Coming to see another actor as trustworthy normally requires the ability to look back upon a lengthy relationship history between the actors, free from disappointments, and breaches of trust (Meyerson et al., 1996). Furthermore, the effect of the length of relationship on trust may be stronger on the interpersonal level. Relationships tend to be stronger on the interpersonal level, driven by shared history and experiences on a personal level (Ma¨kela¨, 2007), whereas longer inter-unit relationships will inevitably involve changes in personnel over time. In addition to this, one may argue that in inter-unit relationships, formal organizational structure may force an MNC unit to continue a relationship with another unit belonging to the same corporation even in the absence of trust, implying that the relation between relationship length and trust might be weaker than between more independent actors (Gulati, 1995) – although past experiences can also lead to a lack of trust within individual network relationships (Ma¨kela¨, 2007). In the light of these considerations, we nevertheless hypothesize a positive relationship between the duration of a relationship – whether individuallevel or unit-level – and the level of trust in that relationship. The following hypotheses are put forth: H1a. The longer the relationship between two managers working in two different units of the same MNC, the higher the level of trust between them.

64

KRISTIINA MA¨KELA¨ ET AL.

H1b. The longer the relationship between two MNC units, the higher the level of trust between them.

2.3. Trust and Communication Frequency Although a certain level of trust is typically viewed as a prerequisite for an actor to be willing to engage in exchange with another, interaction in itself is necessary for such trust to develop between the actors. Initial impressions of trustworthiness may also be reinforced by further interaction, permitting the parties to identify and develop increasing levels of commonalities (Das & Teng, 1998). In a number of previous studies, communication frequency and the level of social interaction have indeed been shown to be positively associated with evaluations of trustworthiness (Tsai & Ghoshal, 1998; Govindarajan & Gupta, 2001; Becerra & Gupta, 2003), therefore facilitating affect-based trust (McAllister, 1995). Frequency being well established as a quality of communication relationships (see, e.g., Monge & Contractor, 2003), it is logical to assume that communication frequency will be positively associated with the level of trust at both interpersonal and inter-unit relationships. Indeed, Kostova and Roth (2003) theoretically propose that the extent of interaction – i.e., the number of contacts and interactions, and the frequency and intensity of these – will be positively related to the social capital of individual ‘boundary spanners’ which will in turn be positively related to the social capital possessed by the MNC units in which they work (Kostova & Roth, 2003). Further, McAllister (1995) found a positive relationship between interpersonal interaction frequency and affect-based trust, and Becerra and Gupta (2003) between interpersonal communication frequency and perceived trustworthiness of the other person. Furthermore, we would expect that, unlike with the length of the relationship which was assumed to have a stronger effect on the interpersonal level, the effect of interaction frequency should be relatively similar on both levels of analysis. Knowledge sharing is a natural product of interaction, and therefore it should follow that the more interaction, the more knowledge sharing on both levels of analysis (Carley, 1991). We consequently advance the following hypotheses: H2a. The higher the communication frequency between two managers working in two different units of the same MNC, the higher the level of trust between them.

Drivers of Interpersonal and Inter-Unit Trust in Multinational Corporations

65

H2b. The higher the communication frequency between two MNC units, the higher the level of trust between them.

2.4. Trust and Cultural Distance The MNC is characterized by multiple internal geographical, cultural, and linguistic boundaries (Westney, 2001), which constrain interaction and may influence the opportunities to build social capital in important ways (Kostova & Roth, 2003). Cultural distance, defined as the degree to which values, norms, and practices differ from one country to another (Kogut & Singh, 1988; Manev & Stevenson, 2001; House, Javidan, Hanges, & Dorfman, 2002) has been suggested in previous research to act as constraint to the efficiency of various transactions within the MNC (Kedia & Bhagat, 1988; Bhagat, Kedia, Harveston, & Triandis, 2002; Leung, Bhagat, Buchan, Erez, & Gibson, 2005). Cultural similarity can be a powerful driver of homophily, i.e., the tendency of similar people to flock together, facilitating trust among the members of the same cultural cluster (Ma¨kela¨, Kalla, & Piekkari, 2007). Inversely, as culture influences ways of thinking, behaving, and communicating (Adler & Graham, 1989; Hofstede, 1980; House, Hanges, Javidan, Dorfman, & Gupta, 2004), national cultural differences may also become a potent ‘source of friction’ (Shenkar, 2001) in interaction among members of different cultures. Furthermore, Manev and Stevenson (2001) found that national cultural distance had a negative impact also on the interpersonal level, i.e., on the strength of interpersonal ties between MNC managers. This effect is, however, likely to be weaker, due to the fact that there is significant intracultural variation among the members of a culture (Au, 2000), driven by personal and professional experiences and backgrounds (Ronen & Shenkar, 1985; Takeuchi, Tesluk, Yun & Lepak, 2005). Taken the above into account, we put forward the following hypotheses: H3a. The higher the national cultural distance between the nationalities of two managers working in two different units within the same MNC, the lower the level of trust between them. H3b. The higher the national cultural distance between two MNC units, the lower the level of trust between them. A graphical summary of our hypotheses is presented in Fig. 1.

KRISTIINA MA¨KELA¨ ET AL.

66

+

H 1a, b. Relationship length

+ H 2a, b. Communication frequency

Trust −

H3a, b. Cultural distance

Fig. 1.

Hypotheses.

3. DATA AND METHODS Our study is based on two data sets collected during 2004–2005, one at the individual level and another at the unit level. While the two sets of data were collected separately, both were obtained from Finnish subsidiaries of foreign MNCs and used the same questions adapted for the two levels of analysis, to achieve a high level of comparability. This research design provided a unique combination of similar data across both interpersonal-level and unitlevel relationships within the MNC. The data collection method used in both data sets was that of structured face-to-face interviews, yielding a high level of validity. The respondents and the researchers went through a pretested questionnaire together. The language in which the questionnaires were administered was English, as this was the language typically used in cross-border intra-MNC interactions. The researchers were prepared to clarify any term respondents might have difficulty in understanding, but this was necessary only in very few instances as the respondents were highly fluent in the English language (the mother tongues of the respondents were Finnish, Swedish, or English). Both the sampling procedures and the two data sets are described in more detail as follows.

3.1. The Individual-Level Data Set The individual level data set consists of 265 observations concerning interpersonal interaction relationships between two managers working in two different units within the same MNC. The data was collected by

Drivers of Interpersonal and Inter-Unit Trust in Multinational Corporations

67

structured interviews with 31 MNC managers working in wholly owned subsidiaries of foreign MNCs located in Finland. The respondents were obtained through a two-stage sampling procedure as a part of a larger research project. In the first phase, the 500 largest companies operating in Finland were identified and this list was grouped into Finnish MNCs and subsidiaries of foreign MNCs. At the second step, the largest MNCs in the list were contacted and individual managers fulfilling the criteria of being involved in frequent internal cross-border interaction were identified. A maximum of three interviewees were sought in one MNC to avoid company bias. Through this procedure, we contacted 59 individuals in total, of whom no one declined but two later cancelled the interview due to pressing work priorities. For the purpose of this analysis, we chose only those individuals who worked for the Finnish subsidiary of a foreign MNC. Each respondent was asked to identify up to 12 colleagues abroad with whom he or she had been in interaction during the previous 12 months, using the following name generator question: ‘‘Think about all your colleagues who work within your company but outside your country. I would like you to indicate three colleagues with whom you have interacted during the last 12 months through each of the following means [four interaction contexts given: non-face-to-face, meetings, project group, crossborder team]’’ (Wasserman & Faust, 1994). This name generator question was designed to provide a maximum variety of relationship contexts, ranging from non-face-to-face, to meeting, project and team contexts, thereby avoiding the problem of only identifying strong relationships, which has been recognized as a typical risk involved in using the name generator technique (Lin, 2001). The respondents were then asked a series of questions concerning the identified relationships. The resulting sample consisted of 265 interpersonal cross-border relationships derived from 31 managers in 23 MNCs (the identities of the individuals and companies are concealed for confidentiality reasons).2 Twelve industries were covered. The relationships bridged 27 countries in five continents. The countries featuring the most individual relationships were Sweden, UK, Denmark, the Netherlands, Switzerland, and the US, and the geographical spread of the examined relationships are summarized in Table 1.

3.2. The Unit-Level Data Set The unit-level data set consists of 102 observations concerning inter-unit relationships, obtained as follows. The data was collected by structured

KRISTIINA MA¨KELA¨ ET AL.

68

Table 1.

Key Sample Characteristics, Individual-Level Data.

Respondents within the sample Relationships within the sample Average number of relationships per respondent Industries within the sample MNCs within the sample Managers within the sample male female Number of countries with relationships to

31 265 8.5 12 23 31 74.3% 25.7% 27

interviews in 61 wholly owned subsidiaries of foreign MNCs located in Finland. We targeted 89 of the 150 largest foreign-owned units in Finland, which were chosen because we had already sampled them once in 2000 for a previous iteration of the same research project. 61 of them agreed. The remaining 28 declined participation either due to lack of time on behalf of the intended respondents or due to organizational changes as compared to the original sampling frame. Data collection took the form of structured interviews with subsidiary top managers, i.e., people in positions such as CEO or country manager. The respondents were asked a series of questions concerning the relationship of the focal subsidiary with (i) the unit’s headquarters and (ii) other units within the MNC (such as a unit in the Nordic countries, a unit in some other part of Europe or a non-European unit). The data collection yielded information on a total of 130 bilateral relations between the focal Finnish subsidiaries and other units belonging to the same parent MNC. However, this included a number of relations to units in countries for which GLOBE data on cultural distance were not available (see the operationalization of cultural distance later) and some cases of missing data. These were excluded, resulting in a final sample of 53 subsidiaries with 102 bilateral relations to headquarters and sister units elsewhere in the world. Statistics on this sample are presented in Table 2.

3.3. Measures The measures used are detailed later. All variables were standardized in order to further assist full comparability between the two data sets.

Drivers of Interpersonal and Inter-Unit Trust in Multinational Corporations

Table 2.

69

Key Sample Characteristics, Unit-Level Data.

Number of subsidiaries in the sample Average number of employees/subsidiary Average annual sales/subsidiary Parent MNC nationality

Number of inter-unit relations

53 555 (SD ¼ 1,148) 127 million US dollars (SD ¼ 191) Nordic n ¼ 16 (30% of sample) European n ¼ 19 (36% of sample) U.S. n ¼ 18 (34% of sample) 102

3.3.1. Dependent Variable 3.3.1.1. Trust. We followed Tsai and Ghoshal (1998) in using trust as a proxy for relational social capital, with measures accordingly adopted from Tsai and Ghoshal (1998). At the individual level, respondents were asked to respond to the following two questions on 7-point Likert-type scales: (i) I can rely on this colleague without any fear of him or her taking advantage of me, even if the opportunity arises, (ii) I can trust this colleague always keeps the promises he or she makes. For the unit level, we reworded the questions to adapt to the level of analysis. Furthermore, in order to ensure that the unit-level measure captured organizational level relational social capital with a satisfactory Cronbach’s a, we complemented these two questions with two additional items (cf. Cortina, 1993), (iii) people from the two units have a sharing relationship; they both freely share ideas, feelings, and hopes about their operations, and (iv) people from the two units have made considerable emotional investments in their working relationship. The mean of these two items were used as the dependent variable (individual-level a ¼ .811; unit-level a ¼ .718). 3.3.2. Independent Variables 3.3.2.1. Length of relationship. At the individual level, the length of the relationship was first measured in months and then recoded into years to account for the typically shorter time perspective associated with interpersonal work-related relationships. At the unit level, relationship length was operationalized as the number of years since knowledge transfer between the subsidiary and the other unit began. 3.3.2.2. Communication frequency. This variable was measured as the sum of three items covering each of the following means of interaction: (i) e-mail, (ii) telephone, and (iii) face-to-face interaction. Sums were used instead of

70

KRISTIINA MA¨KELA¨ ET AL.

averages because the different forms of communication arguably are complementary. At the individual level, respondents were asked to rate on a 6-point scale (1 ¼ daily, 2 ¼ weekly, 3 ¼ monthly, 4 ¼ 3–4 times a year, 5 ¼ once a year or less, 6 ¼ never) how often they interacted with the other person by, respectively, e-mail, telephone, and face-to-face. This scale, adapted from Ghoshal et al. (1994) and Hansen (1999), was reverse coded for the analysis. At the unit level, respondents were asked to rate on 7-point Likert-type scales (ranging from ‘low’ ¼ 1 to ‘high’ ¼ 7) how often e-mail, telephone, and face-to-face communication, respectively, were used as communication channels to transfer knowledge between their subsidiary and the other unit (the scale being adapted from Gupta & Govindarajan, 2000). 3.3.2.3. Cultural distance. This variable was measured as the cultural difference between the nationalities of the two interaction partners in the individual-level data, and the cultural difference between the nationalities of the two units in the unit-level data. We computed the differences using the Kogut and Singh (1988) cultural distance index on nine cultural dimensions identified by the GLOBE study (House et al., 2002, 2004), which provides scales for 62 societies on nine cultural dimensions.3 We used regression predicted (response bias corrected) scores for societal cultural practices scales (House et al., 2004, pp. 742–744), and a composite index was formed by adapting the Kogut and Singh formula (1988) for the nine GLOBE dimensions. The formula is based on the deviation of each of the nine cultural dimensions associated with the nationality of the interaction partner from the score of the respondent’s nationality. Algebraically, it can be presented as follows.  9  X ðI ij  I ir Þ2 =V i CDrj ¼ 9 i¼1 where CDrj stands for the cultural distance of the jth interaction partner’s nationality from the respondent’s nationality, Iij indicates the GLOBE score for the ith cultural dimension and jth nationality, Iir stands for the GLOBE score for the ith cultural dimension and the nationality of the respondent or unit and Vi is the variation of the scores in the ith dimension (adapted from Kogut & Singh, 1988).4 3.3.3. Control Variables 3.3.3.1. HQ relationship. Previous research in MNC contexts (e.g., BarnerRasmussen & Bjo¨rkman, 2007) suggests that levels of relational social

Drivers of Interpersonal and Inter-Unit Trust in Multinational Corporations

71

capital in inter-unit relationships may be higher in HQ-subsidiary relationships than in relationships between two subsidiaries. We therefore need to control for HQ relationships. At the individual level, the question was worded to ascertain whether the interaction partner was located at headquarters ( ¼ 1) or in another unit belonging to the same parent MNC ( ¼ 0). Similarly, at the unit level we used a dummy variable to indicate whether the communication was taking place with the focal unit’s headquarters ( ¼ 1) or another subsidiary unit belonging to the same parent MNC ( ¼ 0). 3.3.3.2. Geographical distance. Proximity of location may be another important driver of trust in interpersonal relationships (Monge & Contractor, 2003; Williams & O’Reilly, 1998), and inversely, geographical distance may complicate the relational bond. Given our argumentation earlier, it is therefore motivated to control for the possible impact of geographical distance. To measure geographical distance, we used the distance in air miles between the locations of the interaction partners (individual-level data) or units (unit-level data). These distances were obtained from Meridian World Data (www.meridianworlddata.com). The distance measures in air miles were recorded as thousands of air miles. 3.3.3.3. Subsidiary size. For the unit-level data, subsidiary size was used as an additional control variable to ensure that more frequent communication would not simply be a product of a larger number of people being involved in interaction. Subsidiary size was operationalized as the log of the subsidiary’s number of employees.

4. RESULTS The correlations of the studied variables are provided in Table 3, together with descriptive statistics. Although there are some significant correlations between the independent variables, neither data set suffers from multicollinearity, as indicated by the low VIF values. Our hypotheses were tested using ordinary least squares regression analysis. Two separate groups of analyses were performed: one for the individual-level data set and another for the unit-level one. The results of these analyses are presented in Table 4. Both models used to test the hypotheses were statistically significant. Hypothesis 1a concerning the association between interpersonal relationship

72

Table 3.

Means, Standard Deviations and Pearson Correlation Coefficients of the Studied Variables.

Variable

Individual-Level Data Mean 5.1 2.8 8.0 1.0 0.3 1.2 –

2

3

4

5

6

7

Standard 1.4 2.9 2.9 0.6 0.3 1.3 –

– 21 29 07 02 04 –

15 – 06 11 11 23 –

24 12 – 11 04 07 –

10 03 16 – 02 01 –

01 21 00 03 – 08 –

08 05 31 00 11 – –

23 04 05 04 17 12 –

Unit-Level Data Mean

Standard

4.9 18.7 13.9 1.0 0.5 1.0 2.3

1.0 21.1 3.4 0.5 0.5 1.2 0.6

Note: Lower diagonal represents individual-level data, upper diagonal unit-level data. Decimal points omitted from correlation coefficients due to space constraints. po.05; .po.01, two-tailed.

KRISTIINA MA¨KELA¨ ET AL.

Trust Relationship length (years) Communication frequency Cultural distance HQ relationship (1 ¼ yes, 0 ¼ no) Geographical distance (1,000 miles) Subsidiary size (log)

1

Drivers of Interpersonal and Inter-Unit Trust in Multinational Corporations

Table 4.

73

Regression Analyses.

Multiple Regression Analysis Individual-level data

Unit-level data

Relationship length Communication frequency Cultural distance HQ relationship Geographical distance Subsidiary size

.209 .268 .016 .010 .071

.148 .235 .064 .053 .017 .254

R R2 Adjusted R2 F

.355 .126 .109 7.452

.374 .140 .085 2.542

Standardized regression coefficients: po0.1, one-tailed; po0.05; po 0.01; po 0.001 for the hypotheses.

length and relational social capital was strongly supported while the corresponding Hypothesis 1b at the inter-unit level of analysis was only marginally supported (at o.1). This suggests that while the length of the relationship is an important driver of both interpersonal-level and unit-level trust, its effect is – as expected – stronger at the interpersonal level. Hypotheses 2a and 2b both received support at the .05 level of significance. This finding indicates that communication frequency is positively related with trust on both levels of analysis. The data did, however, not support the hypothesized positive relationship between the cultural distance between individuals (Hypothesis 3a) and units (Hypothesis 3b) respectively, and trust. Therefore, Hypotheses 3a and 3b will therefore have to be rejected. Among the control variables, only the unit size was significantly related with trust in the unit-level data set.

5. DISCUSSION In this study we set out to explore factors associated with trust as the manifestation of relational social capital within MNCs, both within interpersonal and inter-unit cross-border relationships. Our results show that the drivers of trust are similar across both levels of analysis. Trust was

74

KRISTIINA MA¨KELA¨ ET AL.

found to be significantly and positively related to the length of the relationship between the two individuals or units, and the frequency of the communication between them. The finding that this holds true across both levels of analysis is consistent with Kostova and Roth’s (2003) model of how social capital emerges within the MNC, and also with Mayer, Davis, & Schoorman’s (1995) proposal that similar mechanisms can explain trust both at the interpersonal and inter-unit levels of analysis (see also Schoorman et al., 2007). However, our empirical results, which are among the very few that address both levels within one study, suggest that the relationship may be stronger at the interpersonal level than the unit one, particularly concerning the length of the relationship. This finding is not surprising as there are likely to be a multitude of additional factors affecting the relational dimensions of social capital between MNC units, such as issues related to resource flows or power relationships between subsidiaries. Also, inter-unit relationships inevitably involve changes in personnel over time, which will introduce a level of discontinuity. On the other hand, interunit relationships involve more diverse and multiplex interaction, as communication takes place through various dynamic combinations across different management level and functions. While we were not able to test these issues within the present research design, they represent a fruitful avenue for further empirical research. Contrary to our hypotheses, cultural distance was not found to be significantly related with trust on either the interpersonal or the inter-unit level. Following most studies of cultural distance in international business research we used the Kogut and Singh (1988) cultural distance index as a measure of cultural distance between the nationalities of the interacting managers or unit. Although we addressed some of the criticism directed to the index by using data GLOBE data rather than Hofstede’s dimensions (see, e.g., Shenkar, 2001), the obvious weakness of these measures is that they are based on data at the average national rather than at the interpersonal or unit level (Au, 2000; Ronen & Shenkar, 1985; Takeuchi et al., 2005). In future research, scholars may be able to collect data on cultural distance at focal levels of analysis, although the challenge remains how these constructs can be conceptualized and operationalized. This study suffers from some limitations which at the same time suggest avenues for future research. First, the samples at both the interpersonal and inter-unit levels were relatively small, a factor that may partly explain why some of the statistical relationships were relatively weak in our study. Second, the study was carried out in one location only, Finland. As there may be systematic cross-cultural differences in how trust develops

Drivers of Interpersonal and Inter-Unit Trust in Multinational Corporations

75

(cf. Schoorman et al., 2007), our findings need to be corroborated in other settings. Third, we only examined the level of trust from one side of the relationship, as a dyadic analysis was not possible with the present data. Fourth, data on inter-unit relational social capital was obtained from one person, the general manager or president of the subsidiary. Although this person is likely to be the best individual source of this information, especially in large subsidiaries he or she may not be knowledgeable about the relationships with other MNC units existing in different parts of the focal subsidiary. Fifth, a longitudinal research design would provide a better test of the direction of causality than the present cross-sectional study. Sixth, future studies may go beyond our analysis by examining aspects of the quality and nature of interpersonal and inter-unit communication. Finally, while the fact that both of our data sets replicated similar questions on two levels within the same context was a key contribution beginning to address multilevel issues, the obvious next major research challenge is to empirically test Kostova and Roth’s (2003) model of the mechanisms through which interpersonal social capital over time may evolve into inter-unit social capital. While there exists some evidence of a positive relationship between interpersonal and inter-unit relational social capital (trust) in buyer–supplier relationships (Zaheer et al., 1998), more research is called for concerning MNC-internal social capital using research designs that better addresses the issue of causality. Such research should ideally be longitudinal with data being collected on social capital at different levels of analysis. The propositions presented by Kostova and Roth (2003) provide an excellent starting point for such research.

6. MANAGERIAL IMPLICATIONS The findings of the study carry noteworthy implications for both individual managers as well as for the MNC as a whole. As the importance of relational social capital has been well recognized in previous research (e.g., Kostova & Roth, 2003; Nahapiet & Ghoshal, 1998; Tsai & Ghoshal, 1998) for knowledge and resource exchange, intellectual capital and value creation within firm – and consequently competitive advantage – its facilitation is key for organizational success. For practicing managers, the main message from our research is that communication frequency and the length of the relationship matter for the relational social capital that exists within MNCs – both at the interpersonal and unit levels. Therefore, ideal types of interpersonal and inter-unit

76

KRISTIINA MA¨KELA¨ ET AL.

cross-border relationships are ones in which communication remains frequent for a period of time. The means of initiating such relationships include corporate meetings and symposia with participants from geographically dispersed units, investments in training and development programs with international participation, cross-national and possibly also cross-functional projects and committees, and short- as well as longterm transfers of individuals between units. All of these provide contexts which create interaction and strong relationships between managers from different MNC units (Evans, Pucik, & Barsoux, 2002; Ma¨kela¨, 2007). Although such practices obviously carry a cost, the benefits in terms of enhanced social capital may clearly exceed these expenses. Furthermore, for interpersonal social capital to not only remain a private good of individuals but also become a public good, it is important that boundary spanners both share their experiences (Kostova & Roth, 2003) and help other members of their units to develop relationships with other MNC units. Not all individuals are likely to be equally adept at building relational social capital in international contexts. Therefore, one central question is how to choose people with the skills and attitudes that increase the likelihood of them being able to play the role of boundary-spanners who engage in frequent communication across borders; this will also help organizations to develop social capital at the unit level. Language and communication skills, social skills, and inter-cultural competence are among the qualities that MNC managers may look for in potentially boundary-spanning individuals. MNCs may also want to invest in the development of such skills for its employees. Another personal characteristic likely to contribute to interpersonal trust is an individual’s propensity to trust (Mayer, Davis, & Schoorman, 1995). Without such a propensity, individuals are unlikely to place themselves in a situation where they take the risk of beginning to interact with a stranger and possibly also expose themselves to the risk of opportunistic behavior on the part of the other person. Individuals with a low propensity to trust others are thus less likely to develop a high level of relational social capital than those with a higher propensity. Finally, and as already suggested, more relational social capital may not always necessarily be better (Portes, 1998). If relationship building has a cost for the individual manager, the cost is multiplied for the organization. The organization must therefore create strong ties (Burt, 1992; Hansen, 1999; Krackhardt, 1992) where it matters most, which is a strategic question for each individual organization.

Drivers of Interpersonal and Inter-Unit Trust in Multinational Corporations

77

NOTES 1. The terms ‘trust’ and ‘relational dimension of social capital’ are used in parallel in the analysis that follows. 2. Relationships to nationalities for which GLOBE data (House et al., 2004; see the operationalization of cultural distance) was unavailable were excluded from the analysis. 3. These cultural dimensions include assertiveness, institutional collectivism, ingroup collectivism, future orientation, gender egalitarianism, humane orientation, performance orientation, power distance, and uncertainty avoidance (House et al., 2004). 4. Other cultural distance measures used in previous research include the Kogut and Singh (1988) distance based on Hofstede’s (1980) four cultural dimensions, and Euclidean distance, also based on Hofstede’s (1980) dimensions (Manev & Stevenson, 2001). The obvious reason for choosing the GLOBE dimensions over Hofstede’s is that while building on Hofstede’s work (Leung et al., 2005), the GLOBE scores are both more recent, as well as cover a larger cross-sectional sample and more aspects of culture, responding to the criticism directed towards Hofstede’s measures (see e.g., Shenkar, 2001). For the sake of rigorousness, we tested the Euclidean distance on the nine GLOBE dimensions for the individual-level data set; this produced similar results as the Kogut and Singh (1988) method.

ACKNOWLEDGMENT We would like to thank Tekes, Finnish Funding Agency for Technology and Innovation, for their valuable funding for this research project.

REFERENCES Adler, N., & Graham, J. L. (1989). Cross-cultural interaction: The international comparison fallacy? Journal of International Business Studies, 20, 515–537. Adler, P. S., & Kwon, S.-W. (2002). Social capital: Prospects for a new concept. Academy of Management Review, 27, 17–40. Au, K. Y. (2000). Intra-cultural variation as another construct of international management: A study based on secondary data of 42 countries. Journal of International Management, 6, 217–238. Barner-Rasmussen, W., & Bjo¨rkman, I. (2007). Language fluency, socialization and inter-unit relationships in Chinese and Finnish subsidiaries. Management and Organization Review, 3(1), 105–128. Becerra, M., & Gupta, A. K. (2003). Perceived trustworthiness with the organization: The moderating impact of communication frequency on trustor and trustee effects. Organization Science, 14, 32–44.

78

KRISTIINA MA¨KELA¨ ET AL.

Bhagat, R. S., Kedia, B. L., Harveston, P. D., & Triandis, H. C. (2002). Cultural variations in the cross-border transfer of organizational knowledge: An integrative framework. Academy of Management Review, 27, 204–221. Bolino, M. C., Turnley, W. H., & Bloodgood, J. M. (2002). Citizenship behaviour and the creation of social capital in organizations. Academy of Management Review, 27(4), 505–522. Bourdieu, P. (1983). The forms of capital. Originally published as O¨konomisches Kapital, kulturelles Kapital, soziales Kapital. In: R. Kreckel (Ed.) Soziale Ungleichheiten. (pp. 183–198). Goettingen, Otto Schartz & Co., Reprinted in Granovetter, M. & Swedberg, R. (2001) The sociology of economic life, Cambridge, MA: Westview Press. Brass, D. J., Galaskiewicz, J., Greve, H. R., & Tsai, W. (2004). Taking stock of networks and organizations: A multilevel perspective. Academy of Management Journal, 47(6), 795–817. Burt, R. S. (1992). Structural holes. Boston: Harvard University Press. Carley, K. (1991). A theory of group stability. American Sociological Review, 56, 331–354. Coleman, J. S. (1988). Social capital in the creation of human capital. American Journal of Sociology, 94, S95–S120. Cortina, J. (1993). What is coefficient alpha? Journal of Applied Psychology, 78, 479–495. Das, T. K., & Teng, B.-S. (1998). Between trust and control: Developing confidence in partner cooperation in alliances. Academy of Management Review, 23(4), 491–512. Doz, Y. L., Santos, J., & Williamson, P. (2001). From global to metanational. Boston: Harvard Business School Press. Evans, P., Pucik, V., & Barsoux, J.-L. (2002). The global challenge: Frameworks for international human resource management. Boston: McGraw-Hill Irwin. Ghoshal, S., Korine, H., & Szulanski, G. (1994). Interunit communication in multinational corporations. Management Science, 40, 96–110. Govindarajan, V., & Gupta, A. K. (2001). Global dominance: Transforming global presence into global competitive advantage. San Francisco: Jossey-Bass. Grant, R. M. (1996). Toward a knowledge-based theory of the firm. Strategic Management Journal, 17(Winter Special Issue), 109–122. Gulati, R. (1995). Does familiarity breed trust? The implications of repeated ties for contractual choice in alliances. Academy of Management Journal, 38(1), 85–112. Gupta, A., & Govindarajan, V. (2000). Knowledge flows within MNCs. Strategic Management Journal, 21, 473–496. Hansen, M. T. (1999). The search-transfer problem: The role of weak ties in sharing knowledge across organization subunits. Administrative Science Quarterly, 44, 82–111. Hofstede, G. (1980). Culture’s consequences: International differences in work-related values. Newbury Park, CA: Sage. House, R., Javidan, M., Hanges, P., & Dorfman, P. (2002). Understanding cultures and implicit leadership theories across the globe: An introduction to project GLOBE. Journal of World Business, 37, 3–10. House, R. J., Hanges, P. J., Javidan, M., Dorfman, P. W., & Gupta, V. (2004). Culture, leadership and organizations: The GLOBE study of 62 societies. Thousand Oaks, CA: Sage. Inkpen, A. C., & Tsang, E. W. K. (2005). Social capital, networks, and knowledge transfer. Academy of Management Review, 30(1), 146–165. Jones, G. R., & George, J. M. (1998). The experience and evolution of trust: Implications for cooperation and teamwork. Academy of Management Review, 23, 531–546.

Drivers of Interpersonal and Inter-Unit Trust in Multinational Corporations

79

Kedia, B. L., & Bhagat, R. S. (1988). Cultural constraints on transfer of technology across nations: Implications for research in international and comparative management. Academy of Management Review, 13, 559–571. Kildruff, M., & Tsai, W. (2003). Social networks and organizations. London: Sage. Kogut, B., & Singh, H. (1988). The effect of national culture on the choice of entry mode. Journal of International Business Studies, 19, 411–432. Kogut, B., & Zander, U. (1992). Knowledge of the firm, combinative capabilities, and the replication of technology. Organization Science, 3, 383–397. Kostova, T., & Roth, K. (2002). Adoption of an organizational practice by subsidiaries of multinational corporations: Institutional and relational effects. Academy of Management Journal, 45(1), 215–233. Kostova, T., & Roth, K. (2003). Social capital in multinational corporations and micro–macro model of its formation. Academy of Management Review, 28(2), 297–313. Krackhardt, D. (1992). The strength of strong ties: The importance of philos in organizations. In: N. Nohria & R. Eccles (Eds), Networks and organizations: Structures, form and action (pp. 216–239). Boston: Harvard Business School Press. Leana, C. R., & Van Buren, H. J. (1999). Organizational social capital and employment practices. Academy of Management Review, 24(3), 538–555. Leung, K., Bhagat, R. S., Buchan, N. R., Erez, M., & Gibson, C. B. (2005). Culture and international business: Recent advances and their implications for future research. Journal of International Business Studies, 36, 357–378. Lin, N. (2001). Social capital: A theory of social structure and action. Cambridge: Cambridge University Press. Ma¨kela¨, K. (2007). Knowledge sharing through expatriate relationships: A social capital perspective. International Studies of Management and Organization, 37(3), 108–125. Ma¨kela¨, K., Kalla, H., & Piekkari, R. (2007). Interpersonal similarity as a driver of knowledge sharing within multinational corporations. International Business Review, 16, 1–22. Manev, I. M., & Stevenson, W. B. (2001). Nationality, cultural distance, and expatriate status: Effects on the managerial network in a multinational enterprise. Journal of International Business Studies, 32, 285–303. Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of Management Review, 20, 709–734. McAllister, D. (1995). Affect- and cognition-based trust as foundations for interpersonal cooperation in organizations. Academy of Management Journal, 38, 24–59. McEvily, B., Perrone, V., & Zaheer, A. (2003). Trust as an organizing principle. Organization Science, 14, 91–103. McFayden, M. A., & Cannella, A. A. (2004). Social capital and knowledge creation. Academy of Management Journal, 47(5), 735–746. Meridien World Data. (2005). Retrieved November 30, 2006 from www.meridianworlddata.com. Meyerson, D., Weick, K. E., & Kramer, R. M. (1996). Swift trust and temporary groups. In: R. M. Kramer & T. R. Tyler (Eds), Trust in organizations: Frontiers of theory and research (pp. 166–195). Thousand Oaks: Sage. Monge, P. R., & Contractor, N. S. (2003). Theories of communication networks. New York: Oxford University Press. Nahapiet, J., & Ghoshal, S. (1998). Social capital, intellectual capital and the organizational advantage. Academy of Management Review, 23, 242–266.

80

KRISTIINA MA¨KELA¨ ET AL.

Oh, H., Labianca, G., & Chung, M.-H. (2006). A multilevel model of group social capital. Academy of Management Review, 31(3), 569–582. Portes, A. (1998). Social capital: Its origins and applications in modern sociology. Annual Review of Sociology, 24, 1–24. Putnam, R. D. (1995). Bowling alone: America’s declining social capital. Journal of Democracy, 6, 65–78. Ronen, S., & Shenkar, O. (1985). Clustering countries on attitudinal dimensions: A review and synthesis. Academy of Management Review, 10, 435–454. Schoorman, F. D., Mayer, R. C., & Davis, J. G. H. (2007). An integrative model of organizational trust: Past, present, and future. Academy of Management Review, 32, 344–354. Shenkar, O. (2001). Cultural distance revisited: Towards a more rigorous conceptualization and measurement of cultural differences. Journal of International Business Studies, 32, 519–535. Takeuchi, R., Tesluk, P. E., Yun, S., & Lepak, D. P. (2005). An integrative view of international experience. Academy of Management Journal, 48, 85–100. Tsai, W., & Ghoshal, S. (1998). Social capital and value creation: The role of intrafirm networks. Academy of Management Journal, 41, 464–476. Uzzi, B. (1997). Social structure and competition in interfirm networks: The paradox of embeddedness. Administrative Science Quarterly, 42(1), 35–67. Uzzi, B., & Lancaster, R. (2003). Relational embeddedness and learning: The case of bank loan managers and their clients. Management Science, 49(4), 383–399. Wasserman, S., & Faust, K. (1994). Social network analysis: Methods and application. New York: Cambridge University Press. Westney, E. (2001). Multinational enterprises and cross-border knowledge creation. In: I. Nonaka & T. Nishiguchi (Eds), Knowledge emergence: Social, technical, and evolutionary dimensions of knowledge creation (pp. 147–175). Oxford: Oxford University Press. Williams, K. Y., & O’Reilly, C., III. (1998). Demography and diversity in organizations: A review of 40 years of research. Research in Organizational Behaviour, 20, 77–140. Zaheer, A., McEvily, B., & Perrone, V. (1998). Does trust matter? Exploring the effects of interorganizational and interpersonal trust on performance. Organization Science, 9, 141–159.

This page intentionally left blank

PART II GOVERNANCE

NORMATIVE CONTROL IN MNCS: A MICRO-POLITICAL CONFLICT PERSPECTIVE Susanne Blazejewski ABSTRACT Purpose – The purpose of the chapter is to introduce an actor-centered conflict perspective into research on multinational company (MNC) coordination. We first develop a theoretical framework of conflictual processes in MNC coordination and then use an empirical study of a German MNC in Japan to illustrate how cultural coordination in MNC subsidiaries triggers conflict processes. Methodology/approach – The chapter integrates conflict theory and models of MNC coordination. The empirical study is based on qualitative data. Findings – Coordination programs in MNC such as cultural integration through shared values lead to substantial conflictual processes. Local actors apply micro-political tactics to resist, delay or adjust coordination instruments developed by MNC headquarters. Originality/value of chapter – The chapter applies conflict theory to MNC coordination issues, a field of research which so far is dominated by contingency approaches.

New Perspectives in International Business Research Progress in International Business Research, Volume 3, 83–111 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1745-8862/doi:10.1016/S1745-8862(08)03005-7

83

84

SUSANNE BLAZEJEWSKI

1. INTRODUCTION In a recent book Kristensen and Zeitlin (2005) relate how as a reaction to a British MNC headquarters’ initiative of reassigning global product mandates, conflict arose between the German and Danish subsidiaries. In this situation, the convener of the Danish unit successfully orchestrated a mutual response among its various local stakeholder groups to counter the MNC strategy. This brief example illustrates two important points concerning MNC coordination: (i) MNC headquarters coordination initiatives such as the re-division of labor in its global system produce considerable tension inside the MNC’s international network and among its diverse local constituents. (ii) This conflict potential is handled by the local actors in their own, locally evolved and culturally and institutionally embedded conflict-handling style. In this particular case, the Danish convener was able to draw from the well-established collaborative relationships between the subunit, its local community and the Danish unions. This chapter argues for the introduction of an actor-centered, micropolitical conflict perspective on MNC coordination that particularly takes into account the conflict generated by MNC coordination instruments as well as the individual actors’ reactions to this conflict on the local level of the MNC’s international subsidiaries. So far, micro-level conflictual effects of MNC coordination mechanisms have not been systematically explored in the extensive literature on MNC coordination and control. In this area of research the level of analysis is typically limited to either a one-sided headquarters’ perspective or to the inter-unit relationship between the MNC headquarters and its international subsidiaries. What actually happens inside the subsidiaries concerned is still unclear. How local actors are affected by the MNC control mechanisms and how they actively adopt, modify, manipulate or resist, e.g., the implementation of globally standardized reporting structures or behavioral routines remains within a black box. This is particularly problematic in view of the fact that the success of the MNC coordination effort ultimately depends on the individual actor’s acceptance, internalization and enactment of the desired behavioral routines. A micro-level analysis of conflicts and political reactions to MNC coordination efforts also allows us to systematically investigate their unintended, often costly but hidden negative side effects inside the international subunits which might backlash and eventually threaten to thwart the original intention of the MNC coordination strategy. The chapter is structured as follows: By briefly reviewing dominant streams of the MNC coordination literature we show that micro-level,

Normative Control in MNCs: A Micro-Political Conflict Perspective

85

conflictual effects of MNC coordination are still an underdeveloped research field. Based on rational choice conflict theory we then introduce a micro-political perspective on conflictual processes in MNC coordination which enables us to understand the dual nature of MNC coordination as both handling and triggering conflict. We use a qualitative empirical study of the cultural control mechanisms applied by a German MNC to its Japanese subsidiary to trace local conflicts and the micro-political responses arising during the implementation processes. Drawing from our theoretical discussion and the results of the empirical study, we develop four research propositions on conflictual effects of MNC coordination in the concluding section of the chapter.

2. THREE RESEARCH PERSPECTIVES ON MNC COORDINATION MNC control and coordination are among the most extensively researched topics in the field of international business. We use both terms, ‘‘MNC control mechanisms’’ and ‘‘MNC coordination mechanisms’’ interchangeably in this chapter. Following Harzing, Sorge, and Paauwe (2002, p. 104), we define corporate control mechanisms as ‘‘instruments used to make sure that all units of the organization strive towards common organizational goals’’. They ‘‘comprise all the mechanisms instituted to tie the operations and decisions within and across the components into a larger whole and establish coherence of meaning and purpose within the larger enterprise’’ (Harzing & Sorge, 2003, p. 190). Harzing (1999) distinguishes between four types of MNC coordination summarized in Table 1 (for similar typologies, see Bartlett & Ghoshal, 1998; Child, 1973; Harzing, 1999; Martinez & Jarillo, 1989).1

Table 1.

Direct/explicit Indirect/implicit

Types of Coordination in MNC. Personal/Cultural (Founded on Social Interaction)

Impersonal/Bureaucratic (Founded on Instrumental Artifacts)

Personal centralized control Control by socialization and networks/cultural control

Bureaucratic formalized control Output control

Source: Harzing (1999, p. 21).

86

SUSANNE BLAZEJEWSKI

Relating to our interests in this chapter – conflictual processes and micropolitics in MNC coordination – we can distinguish among three important research streams: (i) a contingency perspective which focuses on the determinants of MNC coordination programs and implicitly regards coordination as an unproblematic, conflict-free, a-political phenomenon; (ii) an inter-unit perspective assigning the MNCs’ subsidiaries a more active, strategizing role vis-a`-vis the headquarters coordination efforts; and (iii) a recent perspective on MNC coordination through global transfer of organizational practices which explicitly acknowledges the inherent conflictual potential in MNC coordination. None of these approaches conceptualize the micro-level, political processes connected to this conflict potential.

2.1. Contingency Perspective This still dominant stream of research focuses on the investigation of macroand meso-level determinants of different MNC coordination strategies such as MNC country of origin, host country nationality, size of operations, global strategy and (headquarters) administrative heritage (Bartlett & Ghoshal, 1998; Calori, Lubatkin, & Very, 1994; Harzing, 1999; Harzing & Sorge, 2003; Wolf, 1994). Some researchers in addition link the control patterns identified to general MNC performance indicators (Harzing, 1999; Nohria & Ghoshal, 1994; Wolf, 1994). Consequently, it has been suggested by a number of researchers (Bartlett & Ghoshal, 1998; Edstro¨m & Galbraith, 1977; Jaeger, 1983; Martinez & Jarillo, 1989; Nohria & Ghoshal, 1994; Roth & Nigh, 1992; Wolf, 1994) that in order to ensure efficient coordination in the dispersed network of today’s MNCs cultural coordination gains in importance. It is, however, remarkable how little emphasis this literature places on the challenges involved in actually implementing the different coordination mechanisms in the MNC’s diverse international subunits. Largely because of a one-sided headquarters’ perspective, authors tend to overlook potential conflicts and local resistances by subsidiary actors against the MNC’s coordination instruments. The highly diverse cultural and institutional contexts in which the international subsidiaries operate have been shown to shape highly diverse – and potentially incompatible – local patterns of actors’ interests, values and work systems (Hofstede, 2001; Whitley, 1999). A quote from Roth and Nigh (1992) on personal coordination exemplifies this stream’s typical view of MNC coordination as an unproblematic, conflict-free and a-political

Normative Control in MNCs: A Micro-Political Conflict Perspective

87

phenomenon: ‘‘personal integrating mechanisms manage the interdependence interests through socialization processes that result in a shared management perspective by headquarters and the subsidiary. The use of personal integrating mechanisms thereby establishes a collaborative relationship between headquarters and the subsidiary, leading subsidiary management to view their relationship with headquarters as effective’’ (Roth & Nigh, 1992, p. 284).

2.2. Inter-Unit Perspective The second stream of research on MNC coordination recognizes the diversity in national, cultural and institutional environments across the dispersed network of the MNC and, as a result, more clearly acknowledges the conflict potential inherent in headquarters’ coordination programs (Ferner, 2000; Kristensen & Zeitlin, 2005; Marschan-Piekkari, Welch, & Welch, 1996). The relationship between local subsidiaries and the MNC headquarters is described as a battleground in which diverse MNC subunits pursuing their own, often local agenda, rather than simply accept the headquarters control policies, try to safeguard their local organizational interests in inter-unit bargaining processes (Geppert, 2003, p. 326; cf. Kristensen & Zeitlin, 2005; Ferner, 2000). Kristensen and Zeitlin (2005), for instance, discuss the role of subsidiary power in handling the conflict potential arising from MNC coordination efforts. But again the consequences of MNC control programs on the micro-level, i.e., on internal processes and relationships within the respective national subsidiary remain outside the researchers’ perspective. By focusing on the inter-unit level of analysis they implicitly assume coherence on the subsidiary level. As we will demonstrate in our empirical study later, however, actors in local subsidiaries cannot be expected to harmoniously pursue identical interests and strategies in response to MNC coordination instruments affecting their individual working environment.

2.3. Organizational Practices Transfer Perspective This third area of research takes a slightly different perspective. Without explicitly emphasizing their coordinative function, authors focus on MNCs’ attempts to standardize organizational practices across diverse global subunits. In particular, they investigate the institutional and organizational

88

SUSANNE BLAZEJEWSKI

determinants of successful (or less successful) transfer results (BeckerRitterspach, Lange, & Lohr, 2002; Kostova, 1999; Kostova & Roth, 2002; Mohan, 2004; Sharpe, 2001; Gooderham, Nordhaug, & Ringdal, 1999; Saka, 2002; Geppert, 2003). Some of the authors, e.g., Sharpe (2001), explicitly acknowledge the conflict potential inherent in transferring practices developed in one context, such as the MNC headquarters, to another cultural, institutional or organizational context. By focusing on determinants and outcomes of practices transfer, however, they fail to systematically address the potentially conflictual transfer process (Blazejewski, 2005). Again, the actual micro-level processes of perception, interpretation, adaptation, resistance or acceptance and internalization of standardized practices by the local, individual actors remain inside a black box. Recently, authors like Golsorkhi (2004) have criticized the strange ‘‘disembodiment’’ of organizational practices research which, as we have demonstrated, tends to neglect the individual actor and his/her active role in conflictual practices transfer processes. Gooderham et al. (1999) also stress the need to bring back micro-level processes and power relations into organizational practices research, which is currently dominated by institutional theory perspectives. This need for a micro-level process perspective is particularly pressing, according to Staw (1991), when, as in the case of MNC practices transfer, conflicts are likely to arise in organizational processes. Similarly, Harzing et al. (2002) and others (Becker-Ritterspach et al., 2002; Geppert, Matten, & Williams, 2002; Lane, 2000) have demanded a more actor-centered approach in research on MNC control and the globalization of organizational practices. ‘‘Work practices in host countries are’’, according to Geppert et al. (2002, p. 57), ‘‘neither determined by the global policies of MNC HQ, not are they entirely shaped by the societal context’’. Instead, individual actors inside the organization have, depending on their power bases, some degree of strategic choice to actively support, oppose, modify or adopt MNC global coordination initiatives in conflictual processes, even within the same national business system. Consequently, individual interests and political strategies have to be systematically re-introduced into the research on MNC coordination. We suggest that a consequent micro-level conflict perspective enables us to open the black box and to overcome some of the current limitations in the international business literature on MNC coordination. In particular, it allows us to tackle three important questions which so far remain open in the MNC coordination literature: How do individual actors on the local subsidiary level react to MNC coordination efforts? If they perceive the

Normative Control in MNCs: A Micro-Political Conflict Perspective

89

coordination instruments to collide with their interests and preferred behavioral routines, how do local actors handle the arising conflicts inside their local subsidiary context? What are the (unintended) effects of these micro-political processes on the overall MNC coordination effort?

3. DEVELOPING A MICRO-POLITICAL CONFLICT PERSPECTIVE ON MNC COORDINATION 3.1. MNC Coordination in Conflict Research The basic assumption of a conflict perspective is that it fundamentally questions the smooth functioning of headquarters’ coordination programs and particularly the implicit assumption about ‘‘automatic’’ (Alvesson & Willmott, 2002, p. 636) compliance to and internalization of headquarters’ norms and procedures by local actors in the different international subsidiaries. Instead, a conflict perspective emphasizes the disruptive, contested effects of coordination mechanisms among local actors inside and at the boundaries of the MNC’s different subunits, searches for barriers to headquarters’ control efforts and traces local processes of resistance, re-interpretation or re-configuration, appropriation and eventual adoption of coordination instruments. The analysis of conflicts implies a process perspective: apart from the generation and manifestation of conflict from coordination initiatives we need to understand what actors actually do to handle the conflict in order to achieve a solution, a compromise or prevent escalation. The success of the coordination initiative ultimately depends on the outcome of these local conflict-handling processes. So far, to our knowledge, there is no sufficiently systematic investigation of MNC coordination using conflict theory. In fact, conflict approaches have in general only been very scarcely applied to the research of the MNC. Notable exceptions are case study based explorations of conflicts between MNCs and host country governments (Fagre & Wells, 1982; Gladwin & Walter, 1980; Luo, 2004). More recently, authors focus on conflict types and their effects on group cohesion and group performance in international project teams (Ayoko, Ha¨rtel, & Callan, 2002; Chevrier, 2003; Joshi, Labianca, & Caligiuri, 2002; Von Glinow, Shapiro, & Brett, 2004). The wellestablished field of comparative research on conflict-handling styles focuses on a cross-rather than inter-cultural perspective and thus overlooks the challenging situation of the MNC where actors from diverse cultural, institutional and organizational backgrounds need to coordinate their

90

SUSANNE BLAZEJEWSKI

behavior and decision-making processes (Leung & Tjosvold, 1998; Morris et al., 1998; Rahim & Blum, 1994; Tinsley, 1998; Ting-Toomey et al., 1991). More relevant to our interests is the work by Kim and Mauborgne (1993/ 1995) who investigate the effect of procedural justice in MNC strategic planning. Subsidiary top-management compliance to global strategic decisions is shown to increase when they perceive the planning process as having been ‘‘fair’’ (Kim & Mauborgne, 1995). Kim and Mauborgne (1993/ 1995) thus focus their attention on selected – albeit important – criteria (procedural justice) for evaluating conflictual decision-making procedures. The conflict process itself, including active strategizing of subsidiary management against the proposed MNC strategy, or potential conflicts arising inside the subsidiary during implementation of the proposed strategy is, however, not considered in their model.

3.2. Conflict Processes: A Conceptual Framework The conflict perspective proposed here is grounded in rational choice conflict theory (Lindenberg, 1985; Olson, 1965; Kahneman & Tversky, 1984; Coleman, 1990). This approach, highly established for the modeling of conflict processes in political science and economics but rarely applied in the management field, takes an actor-centered, micro-political view (Golsorkhi, 2004; Staw, 1991). It is founded on the premises of methodological individualism, but following Giddens (1984) emphasizes that each actor is embedded in her organizational, social, institutional context (determining her ‘‘scope of action’’) which constraints and at the same time enables individual action. Actors hold different interests which they try to safeguard against colliding interests depending on their perceived power bases and situational available alternatives for action (Pettigrew, 1973; Pfeffer, 1992; Bacharach & Lawler, 2000; Dlugos, Dorow, & Farrell, 1993; March & Simon, 1958; Staw, 1991; Blazejewski & Dorow, 2003). The core assumptions of the theory are summarized by Lindenberg (1985) in his RREEMM-model which posits actors as resourceful (actively searching for alternatives for action), restricted (bounded rationality, limited scope of action and decision-making), expecting (regarding the consequences of action), evaluating (subjective assessment of situations, actions, results) and maximizing (safeguarding their own interests) men. Analytically, we can distinguish between three core elements of the conflict process in rational choice theory: conflict generation, conflict handling and conflict outcomes or aftermath (Blazejewski, 2004; Dlugos

Normative Control in MNCs: A Micro-Political Conflict Perspective

91

et al., 1993; Lewicki, Weiss, & Lewin, 1992). In social exchange situations (Blau, 1964; March & Simon, 1958) conflicts arise when an actor’s pursuance of her interests is frustrated or about to be frustrated (Thomas, 1992; Dlugos et al., 1993). Actors try to safeguard their interests against colliding interests through micro-political action based on their available, relevant power bases and for example, culturally preformed patterns of preferred handling alternatives (their ‘‘local rationality’’ Cyert & March, 1963). We identify three different basic strategies of conflict management (Dlugos et al., 1993; Kolb & Putnam, 1992). (i) Communicative conflict handling: the party tries to modify the opponent’s cognitions, i.e., his interests, preferences and values by using communicative means such as persuasion, information, manipulation or verbal threat. Communicative conflict handling requires the opponent’s acceptance of the cognitive changes intended. (ii) Structural conflict handling: the party’s intervention is directed towards the opponent’s scope of action and requires formal power bases. The actor restricts the opponent’s available alternatives for action by adjusting organizational structures, systems, hierarchies and regulations. In this way the conflict is handled even against the opponent’s will; force does not depend on the opponent’s acceptance. (iii) Third party intervention: each party can also call on a third actor (superiors, mentors, mediators, works council, court, colleagues) who on his part can again either apply communicative or structural means of conflict management. Outcomes of the conflict process include the quality and robustness of the ‘‘solution’’ attained (e.g., a compromise), costs (including consumption of power during conflict handling), actors’ satisfaction (e.g., regarding the conflict processes perceived fairness), organizational changes (as a result of structural handling) and escalation effects when the handling procedures generate new conflict issues or enlarge the conflict by attracting additional parties.

3.3. Dual Effects: MNC Coordination as Handling and Triggering Conflict Applying this conflict perspective to the theoretical analysis of MNC coordination the dual nature of coordination mechanisms becomes evident. Depending on the level of analysis and the point of view MNC coordination mechanisms can either be understood as means of conflict handling or as triggers of conflict. The dominant headquarters-oriented, strategic view of MNC coordination implicitly focuses on the first aspect: coordination mechanisms are seen

92

SUSANNE BLAZEJEWSKI

as adequate instruments to manage the fundamental MNC conflict arising from the basically ‘‘strained or even adversarial’’ subsidiary–headquarters relationship (Bartlett & Ghoshal, 1986, p. 88). MNCs are an inherently conflictual arena (Tjosvold, 1999) in which foreign subsidiaries on the one hand respond to their own local context but on the other hand must also be receptive to MNC integration demands in order to allow for joint exploitation of global market inefficiencies and competitive advantages (Morgan, Kristensen, & Whitley, 2001; Prahalad & Doz, 1987). According to Kim and Mauborgne (1993), the conflict potential between the corporate center and the subunits increases when the MNC pursues a global versus a multi-domestic strategy. Coordination mechanisms are implemented to preventively solve this fundamental conflict, effectively manage the subsidiary–headquarters interdependencies and achieve joint pursuance of the overarching MNC global objectives. The conflict-handling function of MNC coordination has not been systematically explored in the international business literature but, without relating to the specific multi-context situation of the MNC, is an established topic in the field of organization theory and organization structure (Bolman & Deal, 1991; Cyert & March, 1963; Dlugos et al., 1993; Katz & Kahn, 1967; Kieser & Kubicek, 1992). Bolman and Deal, for instance, emphasize that ‘‘the basic function’’ of formal coordination through organizational hierarchies is ‘‘conflict resolution’’ (Bolman & Deal, 1991, p. 199). If we transfer their argumentation to the MNC situation we can easily observe how each of the four basic types of MNC coordination mechanisms function as a conflict-handling device: Personal centralized control, for instance, allows formally empowered hierarchical actors to directly determine other actors’ colliding scopes of action through the chain of command. Bureaucratic control aims at conflict prevention by assigning clear, precise tasks, competences and standard procedures to the diverse subsidiaries in the MNC’s global structure. Output control handles conflicts by prescribing desired results of the international subunits’ activities. This strategy does not preemptively manage conflict by channeling interests and actions into predictable structures and standard behavioral routines but assumes implicitly that actors’ interests will inevitably and automatically converge on predefined, mutual outcomes. Cultural control ideally leads to a situation where potentially disruptive conflicts do no longer arise at all because mutual values, internalized by MNC subunit actors, produce assimilated preferences, interests and behavioral patterns. A mutual MNC organizational culture is expected to preemptively dissolve potential causes for conflict generation that evolve from the heterogeneous national or

Normative Control in MNCs: A Micro-Political Conflict Perspective

93

regional cultures and business systems across MNC subsidiaries. According to Kogut and Zander (1996), a shared identity lowers the costs of communication, facilitates coordination and learning and reduces conflicts in organizations. From the perspective of local actors in the MNC’s international subsidiaries, however, the conflict-handling function of coordination mechanisms might be overlaid or even offset by their conflict triggering effects. Coordination mechanisms enforced by headquarters onto highly diverse cultural, institutional and organizational contexts can create substantial conflict potential inside the local recipient unit and in relationships to its external local stakeholders. By changing organizational rules and regulations, norms, reporting structures or decision-making procedures MNC coordination programs affect local actors’ ability to pursue their own interests and redefine their individual behavioral options. MNC coordination mechanisms also threaten to unhinge the precious balance between the heterogeneous demands from the subsidiary’s local and global environment that has been carefully crafted over time. As Becker-Ritterspach et al. (2002, p. 72) rightly point out, particularly formal MNC coordination mechanisms ‘‘strongly influence the distribution of resources, particularly of decisionmaking power’’. In the diverse local contexts of the MNC global network, however, each of the actors concerned potentially perceives the assault on his scope of action in a different way, depending on his internalized cognitive patterns of legitimization and interpretation. Whereas in one subsidiary actors willingly integrate the MNC’s global practices into their behavioral routines, in other subsidiaries coordination efforts trigger substantial conflict on the local level. Focusing on the conflict triggering side of MNC coordination programs, we can theoretically infer the following potential effects on the local level: Direct control mechanisms like bureaucratic formalized control directly interfere with the individual’s behavioral routines by redefining work processes, formal reporting structures and the chain of command. Based on instrumental artifacts, bureaucratic coordination is connected to highly visible organizational devices such as written handbooks or formally defined production processes. Compliance can be ensured by close supervision and the enforcement of sanctions in case of deviating behavior – as long as this supervision is feasible in the dispersed network of today’s MNCs. Despite the potential advantages in the eyes of the MNC’s headquarters management, bureaucratic coordination can be expected to carry substantial conflict potential when applied to international subsidiaries because it tries to directly change the – highly idiosyncratic – ways work is carried

94

SUSANNE BLAZEJEWSKI

out by local employees. Research on organizational practices shows (e.g. Gooderham et al., 1999) that working routines are in fact tightly embedded in the local institutional, cultural and organizational context in which they originally evolved, so that we need to expect high levels of local resistance in response to MNC bureaucratic coordination efforts. On the contrary, we can assume that output control triggers considerable less conflict on the local level because it only indirectly affects the individual’s interests and routine behavior by setting performance targets which the actor is free to attain in his/her personally preferred way. Conflict by output control might arise, however, regarding the definition of performance criteria and the timeframe in which the specified output needs to be attained. Personal centralized control, in turn, is likely to be associated with higher levels of local conflict. In personal centralized control, the superior, e.g., an expatriate delegated to a foreign subsidiary, is empowered by the organizational hierarchy to directly interfere with a subordinate’s scope of action. The conflict potential rises when expatriates employed in an MNC coordination scheme do not possess the required inter-cultural sensitivity or language skills to avoid misunderstandings, thus creating additional conflicts with their local subsidiary employees. We therefore agree with Becker-Ritterspach’s et al. (2002) conclusion that direct mechanisms of MNC control severely curtail local scopes of action and are therefore associated with high conflict potentials. In their empirical study they argue that ‘‘direct mechanisms, in general, and personal centralized control modes, in particular, seemed to make a clash of different rationalities more likely, as local actors are left with little or no freedom to bring in their own ideas and concepts’’ (Becker-Ritterspach et al., 2002, p. 88). We wish to add, though, that the direct personal interaction in, e.g., expatriate–local employee relations also opens up paths for direct communicative handling of this conflict potential through open dialogue and articulation of diverse interests and behavioral traditions – supposing that both parties are willing and able to engage in such dialogue (Von Glinow et al., 2004). Cultural control in MNCs, when pursued as a deliberate strategy, not as the result of evolutionary convergence, has a similarly ambiguous effect on the conflict level. Although authors from the international business field have long considered control by socialization and informal networks as an effective, key element of corporate transnationalization (Bartlett & Ghoshal, 1998; Jaeger, 1983; Roth & Nigh, 1992), we sustain that from the conflict perspective presented here, the dysfunctional conflict potential triggered by socio-integrative coordination is particularly high and dangerous to the

Normative Control in MNCs: A Micro-Political Conflict Perspective

95

overall functioning of the MNC coordination system. Attempts to change deeply held assumptions and often unconscious patterns of thought and evaluation create, as, e.g., Schein (1989) explains, substantial insecurities, anxieties and mental resistance on the part of the actors affected. Similarly, Alvesson and Willmott (2002, p. 622) argue that cultural control (‘‘identity regulation’’) is particularly likely to ‘‘amplify cynicism, spark dissent or catalyze resistance’’. Northrup (1989) shows, in addition, that highly emotionalized, identity-related conflicts such as those to be expected from MNC coordination by socialization, run a greater danger of becoming intractable. The situation is complicated by the fact, that in order to function effectively, cultural coordination requires voluntary acceptance and enactment of the new norms whereas the other coordination alternatives can be more easily enforced. In particular, coordination by output and bureaucratic procedures involves organizational processes which are directly observable and thus easier to control than the intangible constructs of ‘‘values’’ and ‘‘interests’’. Consequently, they are much more easily linked to organizational incentive and sanctioning systems. Compliance to a new set of values, in turn, is not as simple to monitor. Cultural coordination therefore runs a considerable danger of leading to ‘‘ceremonial adoption’’ (Kostova & Roth, 2002, p. 220) where members of the organization espouse the mutual corporate goals and values but do not internalize them (Welch & Welch, 1997). Underlying conflicts between the local actor’s own, personal set of values and the proclaimed mutual MNC culture thus do often not even become visible to the headquarters although they effectively inhibit the MNC’s coordination effort. In these cases MNC control mechanisms create substantial hidden conflict potentials (Kolb & Putnam, 1992) on the local level which eventually offset the MNC’s original coordinative intention. Inconsistencies between the different coordination instruments as applied to the local subsidiary additionally increase the conflict potential. Our theoretical analysis thus suggests that MNC management needs to expect that coordination programs generate considerable conflict potential at the local level even though they are designed and implemented to reduce inter-unit conflict. Based on the theoretical framework introduced earlier, the empirical study in the following section traces how local actors handle the potential conflict generated by global coordination mechanisms. Our case is the Japanese subsidiary of a German MNC. We consider the effects these local handling processes have on the eventual compliance, adoption, adjustment and acceptance of the headquarters’ coordination efforts.

96

SUSANNE BLAZEJEWSKI

4. WHEN GLOBAL COORDINATION TRIGGERS LOCAL CONFLICT: TWO CASES AT DAX/1 JAPAN In this section we analyze two incidents, selected from a broader empirical study of recent coordination strategies of German DAX-30 corporations (Auer-Rizzi, Blazejewski, Dorow, & Reber, 2007), to illustrate how at the DAX/1-company coordination mechanisms introduced by headquarters created substantial conflict inside its Japanese subsidiary and how local actors handled this conflict. Due to limited space and the qualitative methodology applied here, we focus our observations on two exploratory indepth cases which, however, can serve as a basis for the development of preliminary propositions regarding conflict in coordination processes in the concluding section. The study concentrates, in addition, on coordination programs initiated by DAX/1 since 2002 which are, as in many other German companies, dominated by efforts to establish shared goals, values and commitment across international subunits. Whereas German MNCs traditionally pursued either a decentralized approach regarding subsidiary control (Geppert et al., 2002; Lane, 2000), or focused on direct personal coordination by expatriation (Harzing, 1999; Harzing et al., 2002), research has disclosed a recent tendency towards indirect control strategies (Ferner & Varul, 1999; Becker-Ritterspach et al., 2002). These typically focus on cultural control2 which, according to our theoretical analysis, is particularly prone to involve substantial conflict. Table 2 indicates, for example, that between 2000 and 2004 alone more than 75 percent of German DAX corporations have invested into the establishment of shared global core values in order to foster a transnational organizational culture across regional subunits. At the same time, many German MNCs established English as a global corporate language to facilitate organization-wide communication, an often cited prerequisite for the development of mutual dialogue and shared understandings beyond national cultural boundaries (Wolf, 1994; Schneider & Barsoux, 1997). Applying the conflict process perspective on MNC coordination developed earlier, our study concentrates on three core questions: In which way do the headquarters’ attempts to develop a system of cultural coordination at DAX/1 by introducing global core values and English as a common language trigger conflict potential at the local level? How do subsidiary actors handle the arising conflicts in their local organizational environment in order to safeguard their own interests? In which way do these micro-political processes affect coordination outcomes as well as interpersonal relationships at both the local and global levels?

Normative Control in MNCs: A Micro-Political Conflict Perspective

97

Table 2. Global Core Value Initiatives in DAX-30 Corporations. Company Name

Global Core Values Formulated in Years

Adidas-Salomon, Allianz, Bayer, Deutsche Telekom, E.ON, Fresenius Medical Care, RWE, SAP, Schering, Volkswagen Altana, BMW, Deutsche Bank, Deutsche Post, Henkel, Metro, MAN BASF, Commerzbank, DaimlerChrysler, Hypovereinsbank Siemens Continental

2003/2004

Deutsche Bo¨rse, Infineon Technologies, Linde, Lufthansa, Mu¨nchener Ru¨ck, ThyssenKrupp, TUI

2001/2002 1999/2000 1997/1998 The corporation has published core values but the date was unavailable The corporation has not yet published a core value catalogue

Source: Corporate websites, annual reports, press reports.

4.1. Methodology We chose a qualitative, in-depth approach because conflict requires close, trustful interaction between researcher and interviewee which lies beyond standardized questionnaire type surveys. In addition, the lack of in-depth qualitative, micro-level studies on MNC coordination is one of the reasons why the conflict potential of MNC control strategies has remained unrecognized in the international business literature (Ferner & Varul, 1999). Between October 2003 and August 2004 we conducted 25 in-depth interviews in the German headquarters (13 interviews) and the Japanese subsidiary (12 interviews) of DAX/1. All interviewees work in middle or upper management positions and volunteered for this study. Interview length ranged from 90 min to 5 h. Each interview was conducted by two researchers (plus an additional translator in two incidents) as recommended by Eisenhardt (1989). It was conducted in the language requested by the interviewee (Japanese, English, German). A guarantee regarding full anonymity was a prerequisite for an open, informative dialogue. The interviews were loosely structured and allowed interviewees to focus on what they felt relevant and important to talk about (Alvesson, 2003). Grounded in the critical incident method (Flanagan, 1951) interviewees were asked to freely elaborate on their personal experience. Data analysis focusing on content and discursive aspects was conducted by a

98

SUSANNE BLAZEJEWSKI

multinational team of researchers from Germany, the US and Japan to ensure a diversity of cultural perspectives. Interim conclusions and interpretations drawn from the data have been discussed with a number of interviewees.

4.2. Incident I: Implementing the MNC Core Value ‘‘Integrity’’ at DAX/1 Japan A corruption case in DAX/1’s US subsidiary in the late 1990s substantially damaged the corporation’s public image and entailed an uncomfortable court case with a close to $500 million fine. In response, headquarters introduced a set of global core values (regarding integrity, innovation, sustainable performance, environmental responsibility, respectful relationships) in order to prevent future occurrences and re-orient employee behavior to universal ethical standards. Such a core value catalogue is one instrument aimed at cultural coordination, i.e., coordination through shared values and practices. Since 2002, each DAX/1 manager has to express his/ her commitment to the global core values by signing a value statement in an annual ritual. In addition, the core values are tied to management appraisal and incentive programs. From the German headquarters perspective, the results of the coordination effort are highly satisfactory. According to our headquarters’ interviewees, there have been no problems or complaints voiced during the process of implementing the value catalogue in the company’s international subsidiaries. Unnoticed by headquarters, however, the introduction of the core value catalogue produced substantial conflict inside the Japanese subsidiary. In fact, the MNC’s effort to establish company-wide ethical standards severely clashed with local behavioral traditions in the Japanese environment, triggering new, unexpected conflict potentials at the local level. Conflict generation. Japanese tradition demands the regular exchange of gifts, often valuable items, between business partners. The ritual is of high symbolic value for Japanese business people. And, in the cultural context of Japan, it is explicitly not perceived as being associated with illegal practices or corruption but as an indispensable and socially highly acceptable gesture of politeness that strengthens the bonds between business partners. As such, the practice is interpreted by the Japanese interviewees as being an important basis of good business relationships and as contributing to local business success. The introduction of the binding core value catalogue, particularly of the ‘‘integrity’’-value which explicitly prohibits any behavior

Normative Control in MNCs: A Micro-Political Conflict Perspective

99

that could be viewed as being corruptive or violating fair trade, therefore severely curtails the Japanese employees’ scope of action and creates a dilemma for the local sales personnel: If the sales personnel adheres to the norms set by the parent company, they risk a conflict with their Japanese customers and other business partners. In fact, one interviewee recalls several instances where the prescribed regulation has led to unpleasant debates and embarrassing situations with clients who perceived the rejection of gifts offered as being rude and as harming the mutual relationship. If a sales manager in turn sticks with the local cultural tradition and his own internalized norms, conflict potential is generated between him and local senior management which is responsible for implementation of the global practice and needs to legitimize its own doings vis-a`-vis the corporate headquarters. Even though the threat is perceived as being quite diffuse, non-compliance can, according to the Japanese interviewees, potentially lead to poor evaluations in management appraisal procedures and thus damage the personal career outlook. Conflict handling. The internal conflict potential triggered by MNC coordination mechanisms and the local frustration level were described by interviewees as being quite high. The local human resource management therefore initiated a local conflict-handling procedure which over the course of one year gradually alleviated the tension between local business partners, Japanese sales personnel and headquarters-oriented subsidiary management. First, the human resource department initiated a series of talks and workshops to identify the exact causes of conflict and to understand the employees’ ideas, attitudes and individual interests regarding the core value transferred to the Japanese organization. As a result, the behavioral standards as determined by the parent company have been locally readjusted through the formulation of a genuinely Japanese interpretation of the value guidelines and detailed amendments regarding acceptability and legitimacy of precisely defined activities in local operations. According to the revised Japanese regulations, gifts offered to customers or suppliers are again acceptable if they are taken from a predefined pool of corporate presents. Gifts can be received when they are below a certain value and not to be brought to personal use of the respective sales person. The Japanese organization has established an annual procedure where through lottery or auction those presents collected over the year are redistributed among all employees. In addition, a local law firm has been engaged to act as a third party consultant and mediator in conflicts relating to the global value catalogue and its application in Japan. Employees who encounter a conflictual situation regarding gift giving or also regarding price fixing issues which are

100

SUSANNE BLAZEJEWSKI

still occasionally discussed among Japanese managers from competing firms, anonymously call the law firm to receive information and advice on their alternatives for action and their potential (legal) consequences. According to an interviewee who himself has used this service, and the human resources director who established the mediation process, this instrument for third party conflict handling on the local level has been ‘‘quite successful’’ with more than 30 incidents solved in this way over a one year period. Conflict outcomes. The conflict triggered by the MNC coordination initiative inside the Japanese subsidiary has thus been – according to our Japanese interviewees – satisfactorily solved by local actors using local resources and according to their own local conflict-handling style. The local re-interpretation of the integrity-value allows for the continuation of the traditional gift giving ritual – despite the headquarters’ attempt to put an end to any such activity. Because the conflict occurred on the subsidiary level and also was handled locally, however, the MNC’s German headquarters is not even aware of this hidden sabotage of its coordination program.

4.3. Incident II: English as a Common Corporate Language DAX/1 has, in addition, recently introduced English as the worldwide company language in order to improve communication between its diverse constituents and, in particular, to facilitate network building across traditional cultural divides. While the headquarters interviewees again report only positive effects of this policy, including cost reductions in internal communications, the initiative has created substantial, as yet unresolved conflict inside the Japanese subsidiary. In this case, the split runs right through the local organization. Conflict generation. In DAX/1 Japan, the English language policy has distinct negative effects among the older employees who, due to a longstanding lack of foreign language teaching in Japanese educational institutions, do not speak sufficient English. While formerly company documents exchanged with headquarters were translated from German into Japanese or vice versa, they are now received and, even worse, need to be answered in English. The MNC policy, in addition, requires that all management meetings and trainings from junior management levels upwards are conducted exclusively in English. Intranet pages are available in English only. With the introduction of the common corporate language, the older generation feels increasingly excluded from the MNC communication

Normative Control in MNCs: A Micro-Political Conflict Perspective

101

network. Without sufficient language skills many Japanese senior managers cannot be sent to international management trainings or into international project teams. In addition, non-English speakers among the interviewees feel that they can no longer fully participate in many management meetings now conducted in English. According to the interviewees, with English as the corporate language the danger of information asymmetries, misunderstandings and imprecise information exchanged increases in both oral and written communication while at the same time the need for closer interaction with international colleagues inside the global business unit has grown sharply. One manager describes how emails exchanged in rather poor English between the Japanese and the French subsidiary evolved into a full-fledged war fought with ‘‘verbal missiles’’ through recurrent misunderstandings and information failures. Inside the local organization, the privileging of English creates an unhealthy barrier between German expatriate management in the Japanese subsidiary and local, operative high performers who predominantly belong to the older, language-wise less well-versed employees. In contrast, some of the young, rather inexperienced junior managers considerably enhance their position within the internal power structure. Their English skills open them direct communication paths to the German expatriates, who often do not speak Japanese and depend on translation or English as a common language, and to other international managers within the MNC’s global system. Increasingly, information coming from the parent company or global business unit and information going up into the system from the organization’s many non-English speaking managers or operative staff pass through this filter. Language skills and with them an information surplus become new important power bases in the Japanese organization supplanting the more traditional fundament of power based on seniority, formal position and technological expertise. As discussed by Marschan-Piekkari, Welch, and Welch (1999) language in MNCs is clearly related to the internal power structure. The problem aggravates because the Japanese junior managers, functioning as language relays, in the eyes of the older Japanese managers interviewed are suspect of using their information monopoly to their own, personal benefit. They also do neither possess the necessary competence and technical expertise required in their new influential position nor the trust and acceptance of their Japanese colleagues. As a result, there is a general sense within the organization that neither the parent company nor the local expatriate management ‘‘do really know about the things going on in the firm’’. The growing internal split between those who speak English and

102

SUSANNE BLAZEJEWSKI

those who do not already becomes apparent in the interviewees’ use of an ‘‘us/them’’-dichotomy to denote the emerging language-based barriers inside the organization. In addition to the general dissatisfaction and internal tension, some actors at the Japanese subsidiary bluntly reject the implementation of the common language policy. The Japanese human resource manager himself admits that he intentionally refrains from adjusting the local recruitment criteria to prioritize English language requirements as demanded by the German parent company. In his view, this policy would not only increase the problematic generational split inside the organization but would, as he acknowledges during our interview, in the long run threaten his own personal power position since he himself belongs to the group of nonEnglish speakers. The example demonstrates how a single gatekeeper can effectively block the implementation of MNC coordination in order to secure his own, individual interests. Conflict handling. The handling of the internal conflict is characterized by avoidance and below-the-line venting of frustration among Japanese colleagues. Neither the interviewees at the German headquarter nor the expatriates in Tokyo are even aware of the conflict potential triggered by the common language policy. The Japanese interviewees prefer to keep their language-related problems to themselves and thus prevent any kind of ‘‘official’’, coordinated and more efficient conflict handling. Most of them find it too embarrassing to admit that they do not possess the necessary competencies in English to fully participate in the corporate discourse and, apart from some informal complaining among peers, remain silent on the topic. Whenever possible they avoid situations where their lack of English language skills might become noticeable such as international meetings, project groups and training programs. Over the last few years, the Japanese subsidiary has sent only two middle managers into international trainings, simply because of the deficit of adequate language capabilities. Conflict outcomes. Contrary to the original intention of the MNC coordination initiative, the information flow has not been improved for most actors in the Japanese subsidiary. Rather, information asymmetries and misunderstandings – at least from the perspective of the older generation managers interviewed – are on the rise. Instead of improving access to the informal communication network of the MNC, the implementation of English as a common corporate language has led to the partial exclusion of crucial knowledge carriers from the flow of information. Inside the Japanese subsidiary, the new language policy creates high levels of frustration and puts a strain on inter-personal relations,

Normative Control in MNCs: A Micro-Political Conflict Perspective

103

particularly between management generations. Because the conflict is played out predominantly below-the-line, it remains beyond the reach of an efficient, open conflict management. According to Kolb and Putnam (1992), such private, non-confrontational, smoldering conflicts are particularly dangerous to the organization because they tend to produce hidden costs, perpetually damage employee relationships and lead to the institutionalization of inefficient organizational structures. Table 3 presents a summary of the two cases.

Table 3. Coordination Instrument

Conflict Processes Triggered by Headquarters’ Coordination Initiatives at DAX/1-Japan. Headquarters’ Intended results

Conflict Process on Local Level Conflict generation

Global core value ‘‘integrity’’

English language policy

Conflict handling

Conflict outcomes

Re-interpretation of Global core value Communicative value; adoption of handling interferes with locally redefined through local local business dialogue and re- norms - backlash customs, high effect on original interpretation symbolic HQ intention; of core value; meaning of implementation of third party local custom; support structure handling conflict parties: (costs) through outside local sales advisory firm personnel and management, business partners Ceremonial adoption, Communicative Non-English Facilitate subunit forced adoption handling speakers feel coordination with negative side(informal, excluded from through shared effects, high levels below-the-line information language, avoid of frustration, handling), network; conflict through negative effects on conflict conflict parties: misunderstandings internal avoidance English and information relationships, speaking junior asymmetries in change of local employees, global system power structure, older information generation asymmetry employees, HR backlash effect on manager original HQ intention

Avoid conflict between subunits’ local/ global interests through shared values, goals, and commitment

104

SUSANNE BLAZEJEWSKI

5. DISCUSSION AND CONCLUSIONS Both incidents illustrate how conflictual situations on the local subsidiary level are caused by MNC coordination policies which, from the headquarters perspective, were originally intended to actually reduce conflict within the corporation’s international network. In both cases the headquarters remained largely ignorant of the fact that their global coordination efforts triggered internal conflictual processes in the Japanese subsidiary because the local conflict never became manifest on an inter-unit level. As the empirical study shows, however, it is not only necessary to understand how conflict is generated by MNC coordination instruments on the local level. We must also better understand how these conflicts are handled inside the local organizations. The local actors’ way of handling the accruing conflict potential more or less efficiently in the end determines the overall outcome of the MNC coordination policy. In our cases local conflict-handling procedures produced in fact different results: In incident I local conflict handling was initiated by the Japanese human resource department. It supported a structured local dialogue on the conflict arising between the MNC’s global core value catalogue and traditional local business rationalities which eventually led to the reformulation of the critical core value ‘‘integrity’’. In addition, local management established an independent third party advisory structure which helps to resolve ongoing problems of Japanese staff with the MNC’s global culture initiative. As a result, the Japanese subsidiary was able to report the successful implementation of the core value catalogue as demanded by the German MNC headquarters. That the core values, however, had been substantially reinterpreted by the Japanese actors over the course of the local conflicthandling processes in order to allow for continuity of their traditional gift giving rituals never registered at the German parent company – although this re-interpretation basically offsets the MNC’s original intention and thus puts into question the headquarters’ whole idea of creating a common MNC-wide corporate culture – ‘‘coherence of meaning’’ (Harzing & Sorge, 2003) – through coordinative mechanisms. In incident II the local conflict arising due to the MNC’s common language policy is associated with several diverse handling strategies: whereas some actors use their formal power bases, e.g., as gate keeper of recruiting, to actively obstruct the implementation of the MNC policy at the Japanese subsidiary, others take advantage of their English language skills to improve their personal position in the internal power structure. Avoidance is the third dominating handling alternative employed at the

Normative Control in MNCs: A Micro-Political Conflict Perspective

105

Japanese subsidiary by those actors who lack the required language competencies. The immediate results of these micro-level conflict processes include a high level of frustration among the older generation employees, their effective exclusion from many important communication channels in the MNC, the reshuffling of the local power structure in favor of young English speaking but otherwise rather inexperienced junior managers and a consequential blockage of the associated MNC’s recruiting program at the Japanese subunit. These local conflict outcomes produce a backlash effect: instead of improving MNC-wide communication and informal network building as anticipated by the German parent company, the English language policy has led to unintended information asymmetries and harmful internal power games. Again, as in incident I, local conflict processes basically unhinge the MNC’s coordination strategy, and again headquarters remain fully ignorant of this problem. Based on these empirical findings and our theoretical analysis we now summarize our results in four preliminary propositions which can serve as a basis for further research on conflict processes in MNC coordination. Proposition 1. Within the complex organizational, cultural and institutional context of the MNC the implementation of coordination mechanisms by headquarters produces substantial conflict potential on the local subunit level. The conflict potential varies depending on the respective subsidiary, its specific cultural, institutional and organizational environment and local stakeholders’ interests. It is reduced when local interests are preemptively integrated into the MNC coordination program. Proposition 2. The local conflict potential triggered differs depending on the coordination mechanisms applied: it is highest with cultural coordination because it affects deeply held, internalized cognitive and behavioral patterns. It is lowest with coordination by output control which allows for locally idiosyncratic ways of achieving predefined outputs. Proposition 3. How local conflict potential triggered by MNC coordination is handled, depends on actors’ local power bases, their cultural preferences for specific conflict-handling styles, available and accepted conflict-handling procedures offered by the organization (e.g., dialogue platforms, ombudsmen) or the local institutional environment (law, trials, co-determination).

106

SUSANNE BLAZEJEWSKI

Proposition 4. Outcomes of the conflict process differ according to the local actors’ chosen conflict-handling processes. Outcomes potentially include the re-interpretation and modification of the coordination instruments as well as (unintended) negative side effects which might offset the headquarters’ initial coordination strategy. The ‘‘success’’ of MNC coordination programs (implementation and internalization of intended behavior) thus ultimately depends on the outcomes of the local conflict processes. We suggest that the conflict framework introduced in this chapter is constructive in uncovering the consequential but often unintentional and mostly unobserved conflict effects of MNC coordination. With its clear process perspective, it also allows for a better understanding of the intricate linkages between conflict generation, conflict handling and conflict outcomes in the complex environment of the MNC. It thus serves several important functions, e.g., as an analytical perspective for the systematic uncovering of conflict potentials arising from MNC coordination efforts, and as a framework for the development of testable hypothesis for further research on conflict processes in MNC coordination. In addition, it helps us through its comprehensive conceptualization of the conflict-handling alternatives to identify potential starting points for a more efficient management of such conflicts by both local subsidiary and MNC headquarters actors. In our view, conflictual processes as an effect of MNC coordination initiatives are largely unavoidable. In the dispersed network of the MNC local actors in diverse cultural, institutional and organizational contexts continue to develop and pursue their own interests, behavioral routines and local rationalities which then clash with rationalities, patterns of meaning and behavior imposed by the geographically and culturally often distant MNC headquarters. The key aim of an improved practice of MNC coordination therefore needs to be an improved management of the inevitable conflicts arising over the course of the implementation process on the local level. In order to avoid the dysfunctional outcomes associated with hidden, unobserved conflict processes at the local subunit identified in our empirical study, MNC headquarters should in accordance with its international subsidiaries consider the establishment of joint dialogue forums or grievance structures where local actors can voice concerns regarding the MNC coordination programs and indicate their unintentional negative side effects. These channels for more effective, structured conflict handling, however, also need to be open for sensitive topics, e.g., through

Normative Control in MNCs: A Micro-Political Conflict Perspective

107

ensuring the full anonymity of grievance procedures. Our study has demonstrated that local conflict processes triggered by MNC coordination initiatives are likely to involve taboo issues such as the lack of English language skills among Japanese employees.

NOTES 1. The separation between direct/indirect/explicit/implicit control has recently been criticized by Macduffie (2005) and Poppo and Zenger (2002) who – albeit with regard to strategic alliances coordination – emphasize that formal and informal mechanisms frequently interact (‘‘hybrids’’): Informal coordination (e.g., spontaneous information sharing) often comes about only on the basis of pre-established formal structures (e.g., hierarchies, formalized meeting structures). In turn, formal mechanisms can only ‘‘act through’’ (Grandori, 2005) informal mechanisms (e.g., shared understandings) which determine the formal procedure’s interpretation and application. 2. As is to be expected of German bureaucratic home country culture, German MNCs, including DAX/1, tend to apply formal procedures (formulated core values, written rules, explicit incentive and monitoring systems) in their efforts to create a shared culture across international subunits (Auer-Rizzi et al., 2007). This supports our argument above (see note 2) regarding the interdependencies between different types of control mechanisms.

REFERENCES Alvesson, M. (2003). Beyond neopositivists, romantics, and loyalists: A reflexive approach to interviews in organizational research. Academy of Management Review, 28(1), 13–33. Alvesson, M., & Willmott, H. (2002). Identity regulation as organizational control: Producing the appropriate individual. Journal of Management Studies, 39(5), 619–644. Auer-Rizzi, W., Blazejewski, S., Dorow, W., & Reber, G. (Eds). (2007). Unternehmenskulturen in globaler Interaktion. Analysen, Erfahrungen, Lo¨sungsansa¨tze. Wiesbaden: Gabler. Ayoko, O. B., Ha¨rtel, C. E. J., & Callan, V. J. (2002). Resolving the puzzle of productive and destructive conflict in culturally heterogeneous workgroups: A communication accommodation theory approach. The International Journal of Conflict Management, 13(2), 165–195. Bacharach, S. B., & Lawler, E. J. (Eds). (2000). Organizational politics. Stamford, CT: JAI Press. Bartlett, C. A., & Ghoshal, S. (1986). Tap your subsidiaries for global research. Harvard Business Review, November/December, 87–94. Bartlett, C. A., & Ghoshal, S. (1998). Managing across borders: The transnational solution. Boston: Harvard Business School Press. Becker-Ritterspach, F., Lange, K., & Lohr, K. (2002). Control mechanisms and patterns of reorganization in MNCs. In: M. Geppert, D. Matten & K. Williams (Eds), Challenges

108

SUSANNE BLAZEJEWSKI

for European management in a global context – experiences from Britain and Germany. Houndmills, Basingstoke: Palgrave. Blau, P. (1964). Exchange and power in social life. New York: Wiley. Blazejewski, S. (2004). Conflict in multinational corporations: An integrative framework. Proceedings of the 2004 EIBA conference, Ljubljana. Blazejewski, S. (2005). Transferring value-infused organizational practices in multinational companies: A conflict perspective. In: M. Geppert & M. Mayer (Eds), Global, local and national practices in multinational companies. Houndsmills, Basingstoke: Palgrave. Blazejewski, S., & Dorow, W. (2003). Managing organizational politics for radical change: The case of Beiersdorf-Lechia S.A., Poznan. Journal of World Business, 38, 204–223. Bolman, C. G., & Deal, T. A. (1991). Reframing organizations. San Francisco: Jossey-Bass. Calori, R., Lubatkin, M., & Very, P. (1994). Control mechanisms in cross-border acquisitions: An international comparison. Organization Studies, 15, 361–380. Chevrier, S. (2003). Cross-cultural management in multinational project groups. Journal of World Business, 38, 141–149. Child, J. (1973). Strategies of control and organization behavior. Administrative Science Quarterly, 18, 1–17. Coleman, J. S. (1990). The foundations of social theory. Cambridge: Harvard University Press. Cyert, R. M., & March, J. G. (1963). A behavioral theory of the firm. Englewoods Cliffs: Prentice Hall. Dlugos, G., Dorow, W., & Farrell, D. (Eds). (1993). Organizational politics. Wiesbaden: Gabler. Edstro¨m, A., & Galbraith, J. R. (1977). Transfer of managers as a coordination and control strategy in multinational organizations. Administrative Science Quarterly, 22, 248–263. Eisenhardt, K. (1989). Building theories from case study research. Academy of Management Review, 14(4), 532–550. Fagre, N., & Wells, L. T. (1982). Bargaining power of multinationals and host governments. Journal of International Business Studies, 13(2), 9–23. Ferner, A. (2000). The underpinnings of bureaucratic control systems: HRM in European multinationals. Journal of Management Studies, 37, 521–539. Ferner, A., & Varul, M. Z. (1999). The German way: German multinationals and human resource management. London: Anglo-German Foundation for the Study of Industrial Society. Flanagan, J. C. (1951). The critical incident technique. Psychological Bulletin, 54, 327–358. Geppert, M. (2003). Sensemaking and politics in MNCs: A comparative analysis of vocabularies within the global manufacturing discourse in one industrial sector. Journal of Management Inquiry, 12(4), 312–329. Geppert, M., Matten, D., & Williams, K. (2002). Global change management approaches in MNCs and distinct national trajectories: Britain and Germany compared. In: M. Geppert, D. Matten & K. Williams (Eds), Challenges for European management in a global context – experiences from Britain and Germany. Houndmills, Basingstoke: Palgrave. Giddens, A. (1984). The constitution of society: Outline of the theory of structuration. Cambridge: Polity Press. Gladwin, T. N., & Walter, I. (1980). Multinationals under fire: Lessons in the management of conflict. New York: Wiley. Golsorkhi, D. (2004). Strategic practices: What can methodological individualism teach us? Paper presented at the 2004 EGOS conference, Ljubljana.

Normative Control in MNCs: A Micro-Political Conflict Perspective

109

Gooderham, P. N., Nordhaug, O., & Ringdal, K. (1999). Institutional and relational determinants of organizational practices: Human resources management in European firms. Administrative Science Quarterly, 44, 507–531. Grandori, A. (2005). Symposium: New directions in the study of coordination within and between firms. Paper presented at the Academy of Management Meeting 2005, Honolulu. Harzing, A.-W., & Sorge, A. (2003). The relative impact of country of origin and universal contingencies on internationalization strategies and corporate control in multinational enterprises. Worldwide and European perspectives. Organization Studies, 24(2), 187–214. Harzing, A.-W., Sorge, A., & Paauwe, J. (2002). Headquarters–subsidiary relationships in multinational companies: A British–German comparison. In: M. Geppert, D. Matten & K. Williams (Eds), Challenges for European management in a global context – experiences from Britain and Germany. Houndmills, Basingstoke: Palgrave. Harzing, A.-W. K. (1999). Managing the multinationals. Cheltenham: Edward Elgar. Hofstede, G. H. (2001). Cultures consequences: Comparing values, behaviors, institutions and organizations across nations. Thousand Oaks: Sage. Jaeger, A. M. (1983). The transfer of organizational culture overseas: An approach to control in the multinational corporation. Journal of International Business Studies, Fall, 91–114. Joshi, A., Labianca, G., & Caligiuri, P. M. (2002). Getting along long distance: Understanding conflict in a multinational team through network analysis. Journal of World Business, 37, 277–284. Kahneman, D., & Tversky, A. (1984). Choice, values, and frames. American Psychologist, 39, 341–350. Katz, D., & Kahn, R. L. (1967). The social psychology of organizations. New York: Wiley. Kieser, A., & Kubicek, H. (1992). Organisation. Berlin: de Gruyter. Kim, W. C., & Mauborgne, R. A. (1993). Procedural justice, attitudes, and subsidiary top management compliance with multinationals’ corporate strategic decisions. Academy of Management Journal, 36(3), 502–526. Kim, W. C., & Mauborgne, R. A. (1995). A procedural justice model of strategic decision making: Strategy content implications in the multinational. Organization Science, 6(1), 44–61. Kogut, B., & Zander, U. (1996). What firms do? Coordination, identity, and learning. Organization Science, 7(5), 502–518. Kolb, D. M., & Putnam, L. L. (1992). The multiple faces of conflict in organizations. Journal of Organizational Behavior, 13, 311–324. Kostova, T. (1999). Transnational transfer of strategic organizational practices: A contextual perspective. Academy of Management Review, 24(2), 308–324. Kostova, T., & Roth, K. (2002). Adoption of an organizational practice by subsidiaries of multinational corporations: Institutional and relational effects. Academy of Management Journal, 45(1), 213–233. Kristensen, P. H., & Zeitlin, J. (2005). Local players in global games. The strategic constitution of a multinational corporation. Oxford: Oxford University Press. Lane, C. (2000). Globalisation and the German model of capitalism –Erosion or survival? British Journal of Sociology, 51, 207–234. Leung, K., & Tjosvold, D. W. (Eds). (1998). Conflict management in the Asia-Pacific. Singapore: Wiley.

110

SUSANNE BLAZEJEWSKI

Lewicki, R. J., Weiss, S. E., & Lewin, D. (1992). Models of conflict, negotiation and third party intervention: A review and synthesis. Journal of Organizational Behavior, 13, 209–252. Lindenberg, S. (1985). An assessment of the new political economy. Sociological Theory, 3, 99–114. Luo, Y. (2004). A cooptition perspective of MNC–host government relations. Journal of International Management, 10, 431–451. Macduffie, J. P. (2005). Symposium: New directions in the study of coordination within and between firms. Paper presented at the Academy of Management Meeting 2005, Honolulu. March, J. G., & Simon, H. A. (1958). Organizations. New York: Wiley. Marschan-Piekkari, R., Welch, D., & Welch, L. (1996). Control in less-hierarchical multinationals: The role of personal networks and informal communication. International Business Review, 5, 137–150. Marschan-Piekkari, R., Welch, D., & Welch, L. (1999). In the shadow: The impact if language on structure, power and communication in the multinational. International Business Review, 8, 421–440. Martinez, J. I., & Jarillo, J. C. (1989). The evolution of research on coordination mechanisms in multinational corporations. Journal of International Business Studies, 20(3), 489–514. Mohan, A. (2004). Variation of practices across multiple levels within TNCs. Paper presented at the 2004 EURAM conference, St. Andrews. Morgan, G., Kristensen, P. H., & Whitley, R. (Eds). (2001). The multinational firm: Organizing across institutional and national divides. Oxford: Oxford University Press. Morris, M. W., Williams, K. Y., Leung, K., Larrick, R., Mendoza, M. T., Bhatnagar, D., Li, J., Kondo, M., Luo, J.-L., & Hu, J.-C. (1998). Conflict management style: Accounting for cross-national differences. Journal of International Business Studies, 29(4), 729–748. Nohria, N., & Ghoshal, S. (1994). Differentiated fit and shared values: Alternatives for managing headquarters–subsidiary relations. Strategic Management Journal, 15, 491–502. Northrup, T. A. (1989). The dynamic of identity in personal and social conflict. In: L. Kriesberg, T. A. Northrup & S. J. Thorson (Eds), Intractable conflicts and their transformation. Syracuse: Syracuse University Press. Olson, M. (1965). The logic of collective action. Cambridge: Harvard University Press. Pettigrew, A. (1973). The politics of organizational decision-making. London: Tavistock. Pfeffer, J. (1992). Managing with power: Politics and influence in organizations. Boston: Harvard Business School Press. Poppo, L., & Zenger, T. (2002). Do formal contracts and relational governance function as substitutes or complements? Strategic Management Journal, 23(8), 707–725. Prahalad, C. K., & Doz, Y. L. (1987). The multinational mission: Balancing local demand and global vision. New York: The Free Press. Rahim, M. A., & Blum, A. A. (Eds). (1994). Global perspectives on organizational conflict. Westport: Praeger. Roth, K., & Nigh, D. (1992). The effectiveness of headquarters–subsidiary relationships: The role of coordination, control, and conflict. Journal of Business Research, 25, 277–301. Saka, A. (2002). Institutional limits to the internationalization of work systems: A comparative study of three Japanese multinational companies in the UK. European Journal of Industrial Relations, 8(3), 251–275. Schein, E. H. (1989). Organizational culture and leadership. San Francisco: Jossey-Bass.

Normative Control in MNCs: A Micro-Political Conflict Perspective

111

Schneider, S. C., & Barsoux, J.-L. (1997). Managing across cultures. London: Prentice Hall. Sharpe, D. (2001). Globalization and change: Organizational continuity and change within a Japanese multinational in the UK. In: G. Morgan, P. H. Kristensen & R. Whitley (Eds), The multinational firm: Organizing across institutional and national divides. Oxford: Oxford University Press. Staw, B. M. (1991). Dressing up like an organization: When psychological theories can explain organizational action. Journal of Management, 17(4), 805–819. Thomas, K. W. (1992). Conflict and negotiation processes in organizations. In: M. D. Dunnette & L. M. Hough (Eds), Handbook of industrial, organizational psychology. Palo Alto: Consulting Psychologists Press. Ting-Toomey, S., Gao, G., Trubisky, P., Yang, Z., Kim, H. S., Lin, S.-L., & Nishida, T. (1991). Culture, face maintenance, and styles of handling interpersonal conflict: A study in five cultures. The International Journal of Conflict Management, 2(4), 275–296. Tinsley, C. (1998). Models of conflict resolution in Japanese, German, and American cultures. Journal of Applied Psychology, 83, 316–323. Tjosvold, D. (1999). Bridging East and West to develop new products and trust: Interdependence and interaction between a Hong Kong parent and North American subsidiary. International Journal of Innovation Management, 3(2), 233–252. Von Glinow, M. A., Shapiro, D. L., & Brett, J. M. (2004). Can we talk, and should we? Managing emotional conflict in multicultural teams. Academy of Management Review, 29(4), 578–592. Welch, D., & Welch, L. (1997). Being flexible and accommodating diversity: The challenge for multinational management. European Management Journal, 15(6), 677–685. Whitley, R. (1999). Divergent capitalisms. Oxford: Oxford University Press. Wolf, J. (1994). Internationales Personalmanagement. Wiesbaden: Gabler.

This page intentionally left blank

EXPLORING THE ANTECEDENTS OF RELATIONSHIP COMMITMENT IN AN IMPORT–EXPORT DYAD Diana Weinberg and Abraham Carmeli ABSTRACT Although examples of governance abound in the study of export theory, dyadic relationships (importers and exporters) in import theory have thus far received scant attention in the international business literature. Our study aims to explore how high-quality relationships, manifested by trust, respectful engagement and vitality, augment relationship commitment between importer and exporter, while controlling for years of importing, exporter visits, exporter reputation, substitutes, and industry conditions. Data collected from 97 importing companies show that both trust and respectful engagement had a positive effect on relationship commitment. However, vitality mediated the relationship between respectful engagement and relationship commitment. The findings also indicate that the presence of product substitutes had a significant impact on relationship commitment.

New Perspectives in International Business Research Progress in International Business Research, Volume 3, 113–137 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1745-8862/doi:10.1016/S1745-8862(08)03006-9

113

114

DIANA WEINBERG AND ABRAHAM CARMELI

INTRODUCTION Developing economies such as China and India are now playing an ever more significant role in world economic activities. Estimates indicate that the export share of developing countries jumped to 43 percent in 2006 from 20 percent in 1970 (The New Titans, 2006). Recent forecasts by the International Monetary Fund (IMF) suggest that compared to 2.7 percent annual growth in developed countries, developing economies are likely to enjoy on average an annual growth of 6.8 percent (Callen, 2007). With global borders ever more open, and export–import trade flowing at an expanded rate, we need to better understand how importers and exporters build high-quality work relationships. Building quality work relationships is not only important for value creation, but also critical to avoid hold-up situations and to assure value appropriation. Indeed, while task-related variables such as product quality and price have been considered primordial for the buy-task, more recent studies have emphasized the relationship capabilities and motivational aspects of this form of work relationship (Katsikeas, 1998). In addition, while academic studies have tended to focus on exporters as a key subject of inquiry, their counterparts, the importers, have been largely neglected (Jaffe and Ghymn, 2003). Lye and Hamilton (2000) found that 84 percent of the studies on international exchange have been confined to exporters, with very few taking into account the importer side of the equation, and when so, the importer has been regarded as a passive player. Furthermore, despite research calls to consider importers as more active partners in the creation of international business dyads, studies have mainly addressed questions such as the motivation for overseas sourcing, vendor choice, and import management, with only limited attention directed to the overall process of importing (Liang & Parkhe, 1997). While scholars have directed much effort to the study of dyadic relationships in export theory (Styles & Ambler, 2000; Kim & Frazier, 1997; Nijssen, Douglas, & Calis, 1999), relatively little consideration has been devoted to examining exporter–importer exchange in import theory (Jaffe & Ghymn, 2003; Katsikeas, 1998; Liang & Parkhe, 1997; Leonidou, 1989). Furthermore, subsequent to forming business or work relationships, a key question is how parties in the dyad can sustain and solidify this relationship, and avoid attempts at ‘‘hold-up’’ where one party seeks to appropriate most of the created value. Indeed, these issues are yet to be fully understood and need further attention (Liang & Parkhe, 1997).

Antecedents of Relationship Commitment in an Import–Export Dyad

115

The present study seeks to address these research gaps and to contribute to the literature in several ways. First, this study provides a first attempt in developing and testing a model in which high-quality work relationships between exporter and importer play an important role in explaining variation in relationship commitment (i.e., the degree to which each party is committed to the business relationship). Second, we contribute to international business studies by investigating exporter–importer exchange and by studying this phenomenon from the importer’s point of view. This lens is relatively new to the literature because studies thus far have tended to focus on the exporter’s point of view (Marshall & Boush, 2001; Spekman, 1988; Styles & Ambler, 2000), and has left many unanswered questions about the behavior of importers vis-a`-vis their suppliers (Jaffe & Ghymn, 2003; Liang & Parkhe, 1997). Third, we also contribute to the literature by integrating the concept of high-quality work relationships (Dutton, 2003; Dutton & Heaphy, 2003; Ragins & Dutton, 2007), an emerging construct in organizational theory, and apply it to the international context of the firm. In so doing, we not only enrich our understanding of the exporter–importer exchange in international business studies, but also expand the study on the importance of high-quality work relationships in other research contexts. Finally, this study makes an important contribution to better understanding commitment in a work relationship. Unlike many studies on work commitment, which have often examined commitment to objects such as career, job, organization, and work groups, this study is about relationship commitment. Understanding how parties in a relationship may not only sustain but also enhance their relationships and their commitment to the latter is a critical strategic issue.

THEORETICAL BACKGROUND AND RESEARCH HYPOTHESES Relationship Commitment Relationship commitment may yield significant benefits to organizations in a work relationship (Skarmeas, Katsikeas, & Schlegelmilch, 2002). For example, relationship commitment between an exporting and an importing firm enable both to benefit from each other’s networks and gain priority in accessing both significant inputs and distribution channels. Morgan and Hunt (1994, p. 23) define relationship commitment in the context of ‘‘an exchange partner believing that an ongoing relationship with another is so important as to warrant maximum efforts at maintaining it;

116

DIANA WEINBERG AND ABRAHAM CARMELI

that is, the committed party believes that the relationship is worth working on to ensure that it endures indefinitely.’’ Relationship commitment goes beyond a cost-based or calculative exchange; its underpinnings are often relational (Styles & Ambler, 2000), and manifested in trust and respect. As such, commitment is one of the phases of the multidimensionality of the relational exchange, where importer and exporter meet to exchange economic, communication, and/or emotional resources (Dwyer, Schurr, & Oh, 1987). Because relationship commitment is built up over time, the durability of the relationship is an important factor that needs to be considered when studying commitment. Anderson, Hakansson, and Johanson (1994) refer to this as relationship continuity, in which the growth in the relationship increases the economic and psychological benefits to both partners, as it flourishes and deepens. Drawing on this literature, we define relationship commitment as reflecting a caring, stable, significant, valued, and binding exchange.

HIGH-QUALITY WORK RELATIONSHIPS AND RELATIONSHIP COMMITMENT Organizational scholars have recently pointed to the importance of highquality work relationships as ‘‘a rich new interdisciplinary domain of inquiry that focuses on the generative processes, relational mechanisms, and positive outcomes associated with positive relationships between people at work’’ (Ragins & Dutton, 2007, p. 3), as ‘‘high quality relationships are marked by mutual positive regard, trust and active engagement on both sides . . . while low-quality connections corrode motivation, loyalty and commitment’’ (Dutton, 2003, p. 2). Drawing on this new and promising stream of research, this study attempts to integrate the core assumptions of the theory of high-quality relationships to better understand work relationships between exporter and importer in an international research context. In what follows, we develop our model and hypotheses and argue that trust and respectful engagement are important cultivators that give rise, through relationship vitality, to relationship commitment.

RELATIONSHIP VITALITY AND RELATIONSHIP COMMITMENT The term ‘‘vitality’’ is a complex construct, defined by Merriam Webster as ‘‘physical and mental vigor,’’ whose synonyms range from liveliness,

Antecedents of Relationship Commitment in an Import–Export Dyad

117

peppiness and vibrancy to vivacity, enthusiasm, eagerness, and vivaciousness (Merriam Webster Online Thesaurus, 2008). Vitality denotes energy, aliveness, and full functioning (Ryan & Bernstein, 2004; Ryan & Frederick, 1997). ‘‘A vital person is someone whose aliveness and spirit are expressed not only in personal productivity and activity – such individuals often infectiously energize those with whom they come into contact’’ (Ryan & Bernstein, 2004, p. 273). Thus, vitality implies a state of positive arousal, energy and vigor, and a capacity for further development and growth (Ryan & Bernstein, 2004; Ryan & Frederick, 1997). As increased vitality has been found to be linked with positive social contexts (Ryan & Frederick, 1997), relationships between two parties can either be vital or corrosive (Dutton, 2003; Dutton & Heaphy, 2003). In high-quality relationships between two parties, vital nutrients are transferred both ways; these types of relationships are flexible, strong, and resilient. Furthermore, relationship vitality is likely to result in stronger attachment and a higher willingness to cooperate with each other. In high-quality relationships, where there is a high degree of vitality and aliveness flowing from the connection, the parties in the connection are full functioning and thus can produce outcomes that further strengthen their commitment to sustain and enhance the relationship. When parties experience vitality in their relationships, it enables them to create valued identities. This occurs when parties derive positive meaning about what they do and stand for from the relationship (Bartel & Dutton, 2001; Dutton & Heaphy, 2003). This process also creates support and possibilities for both parties to define their identities and the value of each one of them, thus further strengthening commitment to the work relationship. Hence, we hypothesize that the more a relationship between exporter and importer is marked by vitality, the more commitment to this relationship will be increased: Hypothesis 1. There will be a positive relationship between relationship vitality and relationship commitment. What makes work relationships vital and alive? We consider both trust and respectful engagement as keys to inducing vitality in work relationships that two parties form and establish. In what follows, we reason theoretically why and how trust and respectful engagement between two parties induce vitality in their work relationships.

118

DIANA WEINBERG AND ABRAHAM CARMELI

Interpersonal Trust and Relationship Vitality Trust lies at the foundation of nearly all theories of interpersonal relationships. It is referred to as ‘‘one’s expectations, assumptions, or beliefs about the likelihood that another’s future actions will be beneficial, favorable, or at least not detrimental to one’s interests’’ (Robinson, 1996, p. 576). As such, trust is conceptualized in terms of one’s perceived risk of vulnerability within a relationship (Mayer, Davis, & Schoorman, 1995; Rousseau, Sitkin, Burt, & Camerer, 1998). In relationships that are marked by trust, there is a high level of confidence in a party that his or her vulnerability will not be exploited, and that he or she will not be harmed by the behaviors or actions of the other party (Blau, 1964; Deutsch, 1958; Jones & George, 1998; Kramer, 1999; Robinson, 1996; Zand, 1972). Trust ‘‘involves the juxtaposition of people’s loftiest hopes and aspirations with their deepest worries and fears, it may be the single most important ingredient for the development and maintenance of happy, well-functioning relationships’’ (Simpson, 2007, p. 264). If present in a work situation as the foundation between the two parties, trust continuously enhances feelings of happiness that progressively flourish between the parties. In other words, feelings of trust among individuals involved in a work situation will bring about an environment where confidence in the other’s favorable behavior will energize and vitalize the relationship. As trust exists and continues to grow among the two individuals involved, their behavior will change accordingly, bringing about increased feelings of energy and vitality in their interchange. Atwater and Carmeli (in press) found that high quality leadermember exchange cultivates feelings of energy. In addition, Carmeli and Spreitzer (in press) found that trust facilitates connectivity which in turn augments a sense of thriving. This direct relationship between trust and vitality gives rise to the following hypothesis: Hypothesis 2. There is a positive relationship between interpersonal trust and relationship vitality.

Respectful Engagement and Relationship Vitality Respectful engagement is referred to as interacting in a way that communicates a sense of worth and value (Dutton, 2003). Kahn (1990) refers to personal engagement as an expression of the personal self. How much one engages of oneself depends on the work conditions in which one

Antecedents of Relationship Commitment in an Import–Export Dyad

119

is involved. In addition to survival needs, individuals seek to establish a positive social identity through belonging to social aggregates which enhance their worth and esteem. A key question in academic research is how parties in work relationships engage each other in a respectful way and when this manifestation of highquality relationships (i.e., respectful engagement) is likely to occur. Dutton (2003) outlines the following four strategies for establishing respectful engagement in work relationships: (1) present conveying; (2) affirmation and being genuine; (3) active listening; and (4) supportive communication. By conveying presence, one minimizes distraction while interacting with others so that the person is completely focused on the individual. In addition, affirming and being genuine communicate a sense of seeking for value in the other person and a recognition of the person’s existence. In doing so, one expresses genuine interest in the partner and the relationship, thus treating the time spent together as precious and important. This interest in a person is brought forth by an authentic and active manner of listening with true care and empathy. All these elements lead to supportive communication between parties, facilitating a sense of reciprocity. Another factor that is important in creating this respectful engagement is the act of requesting, not demanding, from the other party during work interactions. This transmits to the other respect for his or her needs. Communicating through specific rather than general terms and making statements that are descriptive rather than evaluative add to the thoughtfulness and reciprocity one feels. Throughout this interaction imbued with respect, a sense of worth and value is conveyed rather directly as the person is engaged in a highly energized relationship. We posit that respectful engagement in work relationships between exporter and importer is likely to induce vitality in the connection. As social individuals in need of others, we try to build and maintain satisfying and enriching connections with others. One way to achieve this is through a sense of respect and engagement of the other. This brings about a flow of positive feelings and energy among the individuals as they get to know each other and continue relating in this way. The feelings that emanate from these positive and energizing encounters with each other will continue to follow in the form of feelings of vitality and well-being, as long as continuous respectful engagement is present. In summary, a high degree of engagement manifests energy and vitality (Carmeli, 2009); therefore, consistent with the above, we posit that: Hypothesis 3. There is a positive relationship between respectful engagement and relationship vitality.

120

DIANA WEINBERG AND ABRAHAM CARMELI

The Mediating Role of Relationship Vitality Mediators establish ‘‘how’’ or ‘‘why’’ one variable predicts or causes an outcome variable (Frazier, Tix, & Barron, 2004), or as Baron and Kenny (1986, p. 1176) describe it: ‘‘a mediator is defined as a variable that explains the relation between a predictor and an outcome.’’ In other words, a mediator accounts for the relation between a predictor and criterion (Baron & Kenny, 1986). In our case, the mediator is relationship vitality, and the predictors are interpersonal trust and respectful engagement, while the outcome variable or criterion is the dyadic member’s commitment to the relationship. We suggest that relationship vitality is an important intervening variable in the link between trust and respectful engagement, and relationship commitment between exporter and importer. As previously explained, in our context of international dyadic relationships (importer–exporter), we believe that vitality is a main ingredient in creating and sustaining a committed relationship. Similarly, we argue that trust and relationship engagement will create propitious circumstances for dyadic vitality to flourish. Therefore, we claim that vitality serves as a mediator between trust and respectful engagement in the creation of relationship commitment in the importer–exporter dyad, in that it explains the ‘‘how’’ or ‘‘why’’ trust and respectful engagement lead to a committed relationship. To demonstrate mediation, the variable vitality must have a strong relation to the predictors (trust and respectful engagement) and to the criterion variable (commitment). Hence, following Baron and Kenny’s (1986, p. 1176) argumentation that: (a) variations in levels of the independent variables (trust and respectful engagement, in our case) significantly account for variations in the presumed mediator (vitality, in our case); (b) variations in the mediator (vitality) significantly account for variations in the dependent variable (commitment, in our case); and (c) when these paths are controlled for, the previously significant relations between the independent and dependent variables are no longer significant. If this last path is zero, there is strong evidence for a single dominant mediator. If it is not, then there is a continuum of mediation occurring, and there may be even multiple mediating factors present. As it will be only under these conditions that meditation can occur, we posit that: Hypothesis 4. Relationship vitality mediates the link between interpersonal trust and relationship commitment.

Antecedents of Relationship Commitment in an Import–Export Dyad

121

Hypothesis 5. Relationship vitality mediates the link between respectful engagement and relationship commitment.

METHOD Participants and Data Collection The respondents for this study were sampled from 150 high-tech industry firms headquartered in Israel. The firms represented highly sophisticated technological industries, such as software, hardware, telecommunications, and defense. Data collection lasted from September 2006 to February 2007. The average year of establishment of the firms was 1977. On average, each company had 880 employees, an average of 30% of imports to total purchases, and seven years of import experience. The average number of yearly exporter visits was three. When the final structured questionnaire was ready to be administered, we contacted the import manager of each company, explained the purpose of the study, and assured full anonymity and confidentiality. To encourage participation, we promised to deliver the final results and conclusions of the study. Out of 150 companies that were contacted, we obtained surveys from 105 companies, a response rate of 70 percent. However, complete and useable data were only available for 97 companies. All statistical analyses were performed on this final dataset of firms (N ¼ 97).

Measures Drawing upon the literature on high-quality connections developed by Baker and Dutton (2006), Dutton (2003), Dutton and Heaphy (2003), we proceeded to construct the measures for our proposed model, which included measurements for trust, respectful engagement, vitality, and relationship commitment. To integrate these conceptualizations into our model in accordance with Hinkin (1995), we started by developing our survey based on previously tested items. We subsequently made modifications after consultation with two focus groups, the first one consisting of five senior professors from various business schools from the four largest institutions in Israel, and the second one comprising 10 graduate students at a large Israeli university. Next, a pre-testing, pilot study run on a small

122

DIANA WEINBERG AND ABRAHAM CARMELI

sample of 20 purchasing department participants from high-tech companies generated no significant changes. The final results were five items for interpersonal trust, five items for respectful engagement, three items for vitality, and five items for relationship commitment. Finally, all measurement items were subjected to factor analyses using the cut-off value of .40 for the criterion to decide whether to remove or retain it (Hinkin, 1995). These are outlined below. Respectful Engagement The importer’s perception of his or her relationship with the exporter as it relates to respectful engagement is a key element in the creation of relationship commitment. This way of relating to the other entails complete availability in times of need, a marked interest in the other, and in the time spent together, as well as an explicit and empathetic communication. Respectful engagement was captured by the following five items adapted for this study from Dutton (2003): ‘‘This exporter is always available in case of need’’; ‘‘This exporter expresses genuine interest in our dealings’’; ‘‘This exporter treats our time together as precious’’; ‘‘This exporter listens to us with empathy’’; and ‘‘This exporter communicates in specific rather than general terms.’’ Responses were ranked on a five-point scale, ranging from 1 ¼ strongly disagree to 5 ¼ strongly agree. The Cronbach’s a for this measurement was .80. Interpersonal Trust This measure assesses the extent to which the exporter–importer relationship is based on trust, and to what extent the exporter is acting towards the importer in a way that conveys the latter’s belief in the exporter’s integrity, dependability, and good motives (Dutton, 2003). The five question items for this measurement were: ‘‘This exporter has high integrity’’; ‘‘This exporter treats me in a consistent and predictable fashion’’; ‘‘This exporter is not always honest and truthful’’; ‘‘This exporter’s motives and intentions are good’’; ‘‘I am not fully sure I trust my exporter.’’ Responses were ranked on a five-point scale, ranging from 1 ¼ strongly disagree, to 5 ¼ strongly agree. The Cronbach’s a for this measurement was .80. The measurement items for respectful engagement and interpersonal trust were subjected to a principal components factor analysis with Varimax rotation. The results of this analysis, shown in Table 1, produced two factors that together explain 56.55 percent of the overall item variance. The first factor for respectful engagement (Eigenvalue ¼ 2.91) had factor loadings ranging from .65 to .78, while the second factor for trust

123

Antecedents of Relationship Commitment in an Import–Export Dyad

Table 1.

Factor Analysis Results for Respectful Engagement and Trust. Respectful Engagement (a ¼ .80)

This exporter is always available in case of need This exporter expresses genuine interest in our dealings This exporter treats our time together as precious This exporter listens with empathy This exporter communicates in specific rather than general terms This exporter has high integrity This exporter treats me in a consistent and predictable fashion. This exporter is not always honest and truthful This exporter’s motives and intentions are good I am not sure I fully trust my exporter Percentage of variance explained Eigenvalues

Trust (a ¼ .80)

.65 .76 .78 .75 .70

.13 .11 .20 .14 .05

.30 .34

.74 .60

.00 .19 .01

.74 .71 .83

29.08 2.91

27.47 2.75

(Eigenvalue ¼ 2.75) had factor loadings ranging from .60 to .83. None of the items showed evidence of cross-loadings. Relationship Vitality This measure assesses the extent to which the relationship between the exporter and importer is vital and alive, has a high level of energy and is fully functioning (Dutton & Heaphy, 2003; Ryan & Bernstein, 2004; Ryan & Frederick, 1997). The three items capturing this measure were: ‘‘The relationship with this exporter is vital for the organization’’; ‘‘The relationship with this exporter is alive’’; and ‘‘The relationship with this exporter is reinforcing.’’ Responses were ranked on a five-point scale, ranging from 1 ¼ strongly disagree, to 5 ¼ strongly agree. All three items were subjected to a principal components factor analysis with Varimax rotation. The results of this procedure, shown in Table 2, produced a onefactor solution, that together explains 56.97 percent of the overall item variance (Eigenvalue ¼ 1.71). The factor loadings ranged from .69 to .82. The Cronbach’s a for this measure was .78. Relationship Commitment This measure includes relational factors (Styles & Ambler, 2000), in that it assesses the extent to which the exporter is committed to a long-term

124

DIANA WEINBERG AND ABRAHAM CARMELI

relationship with the importer, where caring, stability, and significance are the underpinnings of the exchange. The five items relating to this measure were: ‘‘This exporter shows caring towards the relationship’’; ‘‘This exporter is very stable’’; ‘‘This relationship is of little significance to both sides’’; ‘‘This relationship is of high importance to the firm’’; and ‘‘Both sides are committed to maintain and enhance the relationship.’’ Respondents were asked to assess these items on a five-point scale (ranging from 1 ¼ strongly disagree, to 5 ¼ strongly agree). These too were subjected to a principal components factor analysis with Varimax rotation. The results of this analysis, shown in Table 3, produced a one-factor solution that explains 48.60 percent of the overall item variance (Eigenvalue ¼ 2.43). The factor loadings ranged from .66 to .75. The Cronbach’s a for this measure was .73. Control Variables We also examined several control variables which had been previously tested in academic research for similar import–export situations. These were Table 2.

Factor Analysis Results for Relationship Vitality. Factor Loadings (a ¼ .78)

The relationship with this exporter is vital for the organization The relationship with this exporter is reinforcing The relationship with this exporter is alive Percentage of variance explained Eigenvalues

.693 .817 .750 56.97 1.71

Table 3. Factor Analysis Results for Relationship Commitment. Factor Loadings (a ¼ .73) This exporter shows caring towards the relationship The relationship with this exporter is very stable The relationship with this exporter is of little significance in the relationship The relationship with this exporter is of high importance to my firm Both sides are committed to maintain and enhance the relationship Percentage of variance explained Eigenvalues

.719 .747 .686 .664 .665 48.60 2.43

125

Antecedents of Relationship Commitment in an Import–Export Dyad

years importing (Marshall & Boush, 2001; number of years importing); exporter visits (Ghymn, Liesch, & Mattsson, 1999; number of visits a year); substitutes (Barney, 1997; Barney & Griffin, 1992; availability of substitute products); exporter reputation (Nijssen et al., 1999; McAllister, 1995); and industry type (Barney, 1997; state of the industry). The six items measuring exporter reputation were subjected to a principal components factor analysis with Varimax rotation. The results of this procedure, shown in Table 4, produced a one-factor solution, which explains 60.65 percent of the overall item variance (Eigenvalue ¼ 3.64). The factor loadings ranged from .67 to .84. The Cronbach’s afor this measure was .87. Similarly, the items for industry were subjected to a principal components factor analysis with Varimax rotation. The results of this analysis, presented in Table 5, produced a one-factor solution, explaining 75.56 percent of the overall item variance (Eigenvalue ¼ 1.51). All factor loadings were above .80. The Cronbach’s a for this measure was .65.

Table 4.

Factor Analysis Results for Exporter Reputation. Factor Loadings (a ¼ .87)

This This This This This This

exporter exporter exporter exporter exporter exporter

is dependable has a very good reputation supplies high quality products is a socially responsible firm is a prestigious firm in the market enjoys high credibility in the market

Percentage of variance explained Eigenvalues

Table 5.

.728 .844 .787 .670 .801 .828 60.65 3.64

Factor Analysis Results for Industry Type. Factor Loadings (a ¼ .65)

There is intense rivalry in the high-tech industry in Israel Changes in the high-tech industry in Israel are rapid Percentage of variance explained Eigenvalues

.869 .869 75.56 1.51

126

DIANA WEINBERG AND ABRAHAM CARMELI

Data Analyses To examine and validate the proposed research model, we used factor analyses, Pearson correlations, and multiple regression analyses to test the strength of the variables, as well as a mediation model between the variables. As far as Pearson correlations are concerned, the issue of multicollinearity needs to be addressed, as our sample size is relatively small (N ¼ 97). When the Pearson correlation coefficient between two or more independent variables crosses the 0.80 line, multicollinearity is suspected, while Nunnally (1978) identifies correlations above 0.70 as an indication for multicollinearity. When employing multiple linear regression (a common procedure in social sciences), one must take into account high correlations between independent variables or predictors due to the possibility of multicollinearity. The main concern is that the regression coefficient could be unstable (Bryman & Cramer, 1994). Research proposes several approaches towards this phenomenon. Jaccard, Turrisi, and Wan (1990) claim that unlike complete multicollinearity (e.g., a correlation exceeding 0.95 in which case the variables should be treated as one), a moderate one does not damage the sum of the ordinary least squares. However, a high correlation between two predictors can cause computational errors in standard computer programs. On the other hand, Belsley, Kuh, and Welsch (1980) claim that this is not a statistics problem but a data problem. It seems that the most balanced approach is that of Mason and Perreault (1991, p. 269): ‘‘In theory there are two extremes: complete collinearity and no-collinearity. In practice, the typical data are somewhere in between these two extremes. Therefore collinearity is a matter of degree, in that certain collinearity always exists, and the challenge is to determine at what point it becomes ‘damaging.’’ There are a few ways to diagnose multicollinearity (Belsley et al., 1980, pp. 85–191). The first step is to review the table of coefficients of the independent variables, but this is not sufficient when an independent variable is a linear combination of several predictors (Chatterjee & Yilmaz, 1992, p. 216). We can therefore now address the question whether our research demonstrates multicollinearity. If it does, then to what degree? Does it have any damaging effect? The first step is therefore to review the correlation table. In the variables system, none of the correlations crossed the 0.65 line. Table 6 shows that the highest correlation was 0.65 between relationship commitment and respectful engagement. According to both

Years importing Exporter–importer visits Substitutes Industry rivalry Exporter reputation Interpersonal trust Respectful engagement Vitality Relationship commitment

Means, Standard Deviations, and Correlations.

Mean

Standard Deviation

1

2

3

4

5

6

7

8

9

6.38 3.53 2.96 4.03 4.03 4.16 3.94 3.97 2.78

4.38 3.76 1.26 .76 .65 .60 .60 .71 .58

1.00 .02 .13 .15 .15 .03 .09 .25 .11

1.00 .11 .10 .21 .03 .09 .13 .09

1.00 .06 .02 .03 .09 .23 .17

1.00 .02 .04 .08 .01 .03

1.00 .45 .59 .45 .57

1.00 .36 .25 .45

1.00 .61 .65

1.00 .64

1.00

Note: N ¼ 97; two-tailed test. po.05; po.01; po.001.

Antecedents of Relationship Commitment in an Import–Export Dyad

Table 6.

127

128

DIANA WEINBERG AND ABRAHAM CARMELI

Belsley et al. (1980) and Nunnally (1978), we can say that this information is sufficient in order to eliminate the concern for damaging multicollinearity. In addition, we tested the mediating effects of vitality on the relationship between both interpersonal trust and respectful engagement on relationship commitment. To this end, we followed Baron and Kenny (1986) and a more recent guideline in Kenny, Kashy, and Bolger (1998). A mediation model can be applied when the following three basic conditions are met: (1) a significant relationship between the dependent variable (in our case, relationship commitment) and the independent ones (in our case, interpersonal trust and respectful engagement) is established; (2) a significant relationship between the mediator (in our case, vitality) and the independent variable is established; (3) the significant relationship between the dependent variable and the independent ones becomes non-significant when the mediator is specified in the model. In other words, the mediator will explain the relation or causal mechanisms between a predictor and an outcome (Frazier, Tix, & Barron, 2004). The regression equations which need to be performed to establish mediation will ultimately determine if mediation is in fact present.

RESULTS The means, standard deviations, and correlations among the research variables are presented in Table 6, with additional descriptive statistics shown in Table 7. The bivariate correlations indicate that all independent (respectful engagement; trust) and moderator (vitality) variables were Table 7. Descriptive Statistics of the Research Variables. Scale Cronbach’s a Minimum Maximum Mean

Years importing Exporter–importer visits Substitutes Industry rivalry Exporter reputation Interpersonal trust Respectful engagement Vitality Relationship commitment

W0 0þ 1–5 1–5 1–5 1–5 1–5 1–5 1–5

– – – .65 .87 .80 .80 .78 .73

.50 .00 1.00 2.00 1.00 1.80 1.20 1.67 1.60

20.00 25.00 5.00 5.00 5.00 5.00 5.00 5.00 4.75

6.38 3.53 2.96 4.03 4.03 4.16 3.94 3.97 2.78

Standard Deviation 4.38 3.76 1.26 .76 .65 .60 .60 .71 .58

Antecedents of Relationship Commitment in an Import–Export Dyad

129

significantly related to the dependent variable, relationship commitment (r ¼ .65, po.001; r ¼ .45, po.001; r ¼ .64, po.001, respectively). Nevertheless, although the mediator (vitality) and independent variable of respectful engagement were positively related (r ¼ .61, p ¼ n.s.), results showed that the mediator (vitality) and the other independent variable (interpersonal trust) were not correlated (r ¼ .25, po.001). Moreover, results showed that exporter reputation was significantly related to the independent, moderator and dependent variables (r ¼ .45, po.001; r ¼ .59, po.001; r ¼ .45, po.001; r ¼ .57, po.001 for trust, respectful engagement, vitality and relationship commitment, respectively). However, the other control variables (years importing, exporter–importer visits, substitutes, industry) did not have a significant effect on relationship commitment (all correlations were below .09). All Cronbach’s as for the items were significant and above .65, hence supporting the inter-correlation and the reliability of these variables. As the studies consider these variables as central to their theories (Marshall & Boush, 2001; Ghymn et al., 1999; Ghymn et al., 1999; Barney, 1997; Barney & Griffin, 1992; Nijssen et al., 1999; McAllister, 1995), our study found that they had little or no impact on the commitment generated within the dyad. To test the model’s hypotheses, we conducted a series of Pearson correlations and multiple regression analysis, along with mediation analysis to test and verify the proposed paths. In each progressive regression equation, the control variables were entered as the first step, after which we checked each hypothesis in turn. Mediators establish ‘‘how’’ or ‘‘why’’ one variable predicts or causes an outcome variable. A mediator is defined as a variable that explains the relation between a predictor and an outcome (Baron & Kenny, 1986). In their most recent work on moderation and mediation, Frazier et al. (2004) clearly define and establish the steps required to fit a model of mediation, where a variable mediates the relation between a predictor variable and an outcome variable: (1) a significant relation between predictor and outcome must be present; (2) the predictor is related to the mediator; (3) the mediator must be related to the outcome variable; (4) the strength of the relation between the predictor and the outcome is significantly reduced when the mediator is added to the mediation model. This process implies a causal chain of events, where the mediator is also assumed to be caused by the predictor variable and to cause the outcome variable (Kenny et al., 1998). Hypothesis 1, which posited a positive relationship between relationship vitality and relationship commitment, was supported. As can be seen from the results under Model 3, Table 8, the relationship between vitality and

130

DIANA WEINBERG AND ABRAHAM CARMELI

Table 8. Regression Results for the Relationships between Trust, Respectful Engagement, Vitality, and Relationship Commitment. Model 1

Model 2

Model 3

Model 4

Vitality

Relationship Commitment

Relationship Commitment

Relationship Commitment

Constanta

1.14

1.72

1.00

1.72

Years importing Exporter visits Substitutes Industry rivalry Exporter reputation R2 Adjusted R2 F for R2 Degrees of freedom Vitality DR2 F for DR2 R2 Adjusted R2 Degrees of freedom Interpersonal trust Respectful engagement DR2 F for DR2 R2 Adjusted R2 Degrees of freedom

.13 .01 .26 .04 .16 .29 .25 8.02 5, 99

.02 .04 .27 .00 .08 .21 .17 5.20 5, 99

.01 .50 .17 14.93 .46 .42 2, 97

.37 .28 .19 15.14 .40 .35 2, 97

.11 .07 .15 .00 .21 .21 .17 5.20 5, 99 .45 .14 21.54 .35 .31 1, 98

0.07 0.04 0.18 0.02 0.02 .21 .17 5.20 5, 99 .37 .14 21.54 .35 .31 1, 98 .37 .10 .12 10.87 .47 .43 2, 96

po.05; po.01; po.001. a

Unstandardized coefficients.

relationship commitment was significant and positive in direction (b ¼ .45, po.001). Nevertheless, the results of Model 1, Table 8 reject Hypothesis 2, which posited a positive relationship between interpersonal trust and relationship vitality (b ¼ .01, p ¼ n.s.), but support Hypothesis 3, which posited a positive relationship between respectful engagement and vitality (b ¼ .50, po.001). These first steps in our regression and mediating analysis show that a significant relationship was established between the mediator (vitality) and the independent variable (respectful engagement, although not with trust), and between the mediator (vitality) and the dependent variable

131

Antecedents of Relationship Commitment in an Import–Export Dyad

(relationship commitment). The final step determines whether the significant relationship between the dependent variable (relationship commitment) and the independent one (respectful engagement;) becomes non-significant once the mediator (vitality) is specified in the model. In this way, mediation will be accounted for. The findings of the mediation model are depicted in Fig. 1 and Table 8. Towards this end, we specified the mediator (vitality) in the model (Model 4, Table 8). These show that the path from trust to vitality was not significant (b ¼ .01, p ¼ n.s.), while the path between trust and relationship commitment was significant (b ¼ .37, po.001). This indicates that the relationship between trust and relationship commitment is not mediated by vitality, but rather that there is a direct connection between them, thus rejecting Hypothesis 4, which posited that relationship vitality would mediate the relationship between trust and relationship commitment. On the other hand, Hypothesis 5, which predicted that vitality would mediate the relationship between respectful engagement and relationship commitment, was supported. As can be seen in Fig. 1, and in Models 2, 3, and 4 in Table 8, the connection between respectful engagement and relationship commitment became non-significant when the mediator (vitality) was specified (b ¼ .28, po.01 vs. b ¼ .10, p ¼ n.s.), and the effect of vitality on relationship commitment remained significant (b ¼ .45, po.01 vs. b ¼ .37, po.01), suggesting that vitality mediated the relationship between respectful engagement and relationship commitment.

Years Importing .37***

/

.37*** -.07

Interpersonal Trust

Exporter Substitutes Visits -.04

-.18*

Supplier Reputation

-.02 .02

.01

.50***

Relationship .45*** / .37*** Vitality

Relationship Commitment

Respectful Engagement .28** / .10 N = 97;* p < .05;** p < .01;*** p < .001

Fig. 1.

The Research Model.

Industry

132

DIANA WEINBERG AND ABRAHAM CARMELI

DISCUSSION As an increased globalized workforce continues to answer the economic and business needs of firms, we can no longer avoid the study of importer behavior, a neglected part of international business studies in need of exploration (Liang & Parkhe, 1997). Based on the literature reviewed (Jaffe & Ghymn, 2003; Lye & Hamilton, 2000; Dutton, 2003; Dutton & Heaphy, 2003; Dutton & Ragins, 2007), we believe that the notion of highquality connections will elucidate the ‘‘whys’’ and the ‘‘hows’’ of relationship commitment within the importer–exporter dyad. For this purpose, we developed and tested a model of high-quality interpersonal connections in import theory. In our exploration of the role of these types of connections (trust, respectful engagement, and vitality) in augmenting relationship commitment between importer and exporter, our model showed that both trust and respectful engagement had a positive effect on relationship commitment. In addition, vitality mediated the relationship between respectful engagement and relationship commitment. The only control variable that had a significant impact on relationship commitment was the presence of product substitutes.

Theoretical Implications The starting point of this study was to attempt the development and testing of a model in which Dutton’s (2003) high-quality relationships (trust, respectful engagement, vitality) in an international context (importer– exporter dyad) would explain the variation in relationship commitment between importer and exporter. In this way, the intersection of organizational behavioral and international business was targeted. In addition, in so doing, we aimed to demonstrate the validity and applicability of the concept of high-quality work relationships (Dutton, 2003; Dutton & Heaphy, 2003; Dutton & Ragins, 2007) in an international context. Finally, in studying relationship commitment between individuals, and not commitment to objects (career, job, organization or work group), we hoped to contribute and enrich a different dimension of the concept of interpersonal commitment. To answer these queries, our findings show that, as posited, respectful engagement and interpersonal trust between the members of the dyad are necessary preconditions for the presence of relationship commitment, with vitality as mediator of these variables. The data clearly show how

Antecedents of Relationship Commitment in an Import–Export Dyad

133

vitality mediates respectful engagement, but not trust, into creating commitment to the relationship. Trust has a more direct effect in the creation of commitment and does not necessitate the presence of vitality to affect the relationship. Although previous theory included years importing, exporter visits, industry type and supplier reputation as important variables in the context of an import–export situation (Marshall & Boush, 2001; Ghymn et al., 1999; Ghymn et al., 1999; Barney, 1997; Barney & Griffin, 1992; Nijssen et al., 1999; McAllister, 1995), our study found that they had little or no impact on the commitment generated within the dyad. The only important variable to affect commitment was the absence or presence of product substitutes (Barney, 1997; Barney & Griffin, 1992), in that the threat of a substitute product or exporter may sway the importer into modifying his or her relationship with the existing exporter, thereby affecting his or her commitment to the specific exporter.

Managerial Implications Previous studies have directed efforts to understanding how import–export dyads are formed (Liang & Parkhe, 1997). Nevertheless, the actual workings of the relationship have yet to be addressed (Jaffe & Ghymn, 2003). The proposed model and the results obtained clearly contribute to the theory of international business and organizational behavior, where almost no research between the intersection of these disciplines exists. Our model has shown that for managers trying to create more committed relationships, it is mainly interpersonal connections, in the form of trust, engagement with respect, and vitality that need to be watched and nurtured from the onset. On further examination, our results show that vitality is pivotal in creating commitment in the relationship. How it is created and increased should be the subject of further scientific study. Among the questions purchasing managers should ask themselves when entering into an international exchange are: ‘‘How do I develop feelings of trust, engagement and vitality, which I know will lead to a relationship commitment and to a successful relationship for my exporter and for me?’’ ‘‘How do I enhance the exchange so that it is fruitful, committed and vital, and what characteristics will facilitate this exchange?’’ Further research needs to uncover ways to create the necessary conditions conducive to these goals.

134

DIANA WEINBERG AND ABRAHAM CARMELI

Limitations and Future Research Directions Nevertheless, as with every research study, this one has several limitations and constraints. First, although the data for this study came from firms based in Israel (with exporters abroad), the model cannot be generalized and should be replicated in and extended to different cultural settings to test its validity. Second, because our sample consisted mostly of high-tech companies (albeit including different industries within this category), additional study of other industry types could extend the potential applicability of the model. Third, including other behavioral/socialization characteristics could result in a more complete framework integrating Dutton’s work (2003) on high-quality connections, as little research has been done on the behavioral aspects of import–export theory (importer side). Therefore the ‘‘marriage’’ of international marketing and organizational behavior would take on an added dimension, as there is virtually no research in the intersection of these two disciplines. Fourth, other criteria to expand and enrich the model could measure the commitment to the dyad in financial terms, hence giving additional support to the importance of commitment in a dyadic work relationship. Fifth, another interesting question is whether the behavior of local purchasers and importers vis-a`-vis their suppliers (exporters) differs or not. As there are no data or model to support this inquiry either way, further study into this interesting area of organizational buying behavior is necessary. Finally, it is important also to note that each set of data consisted of one questionnaire per purchasing team of each firm, as most companies had one person in charge of international purchasing. As these self-reports are subjective and can lead to data inflation, further research could contribute from having data from at least two members per purchasing team.

Conclusion Our study was undertaken first and foremost to answer calls in the literature in relation to the behavior of importers, which has so far received scant attention from business scholars. Taking organizational behavior science as our starting point, we applied it to a dyadic importer–exporter relationship in order to uncover what makes the importer ‘‘tick’’ in committing to his/her exporter. Dutton’s (2003) work on high-quality connections provided us with notions such as relationship commitment, interpersonal trust, respectful engagement, and vitality in building our behavioral model.

Antecedents of Relationship Commitment in an Import–Export Dyad

135

Our study’s findings support the notions that interpersonal trust and respectful engagement between importer and exporter are prerequisites in creating a long-term committed business relationship. In addition, our study uncovered the role of vitality as it mediates the effects of respectful engagement on relationship commitment. Although many questions still remain unanswered, the study’s results provide a starting point from which additional research can be undertaken to further uncover how a long-term, committed relationship can be created, and how and why it affects work performance in an international setting both for theory development, and also for managers who wish to understand and develop successful and committed international ventures, imbued with vitality, respect, engagement, and trust among the individuals of the import–export dyad.

REFERENCES Anderson, J. C., Hakansson, H., & Johanson, J. (1994). Dyadic business relationships within a business network context. Journal of Marketing, 58(4), 1–15. Atwater, L., & Carmeli, A. (in press) Leader-member exchange, feelings of energy and involvement in creative work. The Leadership Quarterly. Baker, W., & Dutton, J. E. (2006). Enabling positive social capital in organizations. In: J. E. Dutton & B. Ragins (Eds), Exploring positive relationships at work (pp. 1–35). Mahwah, NJ: Lawrence Erlbaum. Barney, J. B. (1997). Gaining and sustaining competitive advantage. Reading, MA: Addison-Wesley. Barney, J. B., & Griffin, R. W. (1992). Management of organizations. Boston, MA: Houghton Mifflin Company. Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. Bartel, C. A., & Dutton, J. E. (2001). Ambiguous organizational memberships: Constructing social identities in interactions with others. In: M. A. Hogg & D. Terry (Eds), Social identity processes in organizational contexts (pp. 115–130). Philadelphia, PA: Psychology Press. Belsley, D. A., Kuh, E., & Welsch, R. (1980). Regression diagnostics: Identifying influential data and sources of collinearity. New York: Wiley. Blau, P. M. (1964). Exchange and power in social life. New York: Wiley. Bryman, A., & Cramer, D. (1994). Quantitative data analysis for social scientists. London: Routledge. Callen, T. (2007). IMF forecasts slower growth for 2008, IMF Research Department, October 17. Carmeli, A. (2009). High-quality relationships, individual aliveness and vitality, and job performance at work. In: N. Ashkanasy, W. J. Zerbe & C. E. J. Ha¨rtel (Eds), The effect of affect in organizational settings, research on emotion in organizations (Vol. 5). Oxford, UK: Elsevier/JAI Press. Forthcoming.

136

DIANA WEINBERG AND ABRAHAM CARMELI

Carmeli, A., & Spreitzer, G. M. (in press). Trust, connectivity, and thriving: Implications for innovative behaviors at work. Journal of Creative Behavior. Chatterjee, S., & Yilmaz, M. (1992). A review of regression diagnostics for behavioral research. Applied Psychological Measurement, 16(3), 209–227. Deutsch, M. (1958). Trust and suspicion. Journal of Conflict Resolution, 2, 265–279. Dutton, J. E. (2003). Energize your workplace: How to create and sustain high-quality connections at work. University of Michigan Business School Management Series: Wiley. Dutton, J. E., & Heaphy, E. D. (2003). The power of high-quality connections. In: K. S. Cameron, J. E. Dutton & R. E. Quinn (Eds), Positive organizational scholarship (pp. 263–278). Mahwah, NJ: Lawrence Erlbaum. Dutton, J. E., & Ragins, B. R. (2007). Moving forward: Positive relationships at work as a research frontier. In: J. E. Dutton & B. R. Ragins (Eds), Exploring positive relationships at work: Building a theoretical and research foundation (pp. 387–400). Mahwah, NJ: Lawrence Erlbaum. Dwyer, F. R., Schurr, P., & Oh, S. (1987). Developing importer–seller relationships. Journal of Marketing, 52(2), 11–27. Frazier, P. A., Tix, A. P., & Barron, K. E. (2004). Testing moderator and mediator effects in counseling psychology research. Journal of Counseling Psychology, 51(1), 115–134. Ghymn, K., Liesch, P., & Mattsson, J. (1999). Australian import managers’ purchasing decision behavior: An empirical study. International Marketing Review, 16(3), 202–215. Hinkin, T. R. (1995). A review of scale development practices in the study of organizations. Journal of Management (3), 1–23. Jaccard, J., Turrisi, R., & Wan, C. K. (1990). Interaction effects in multiple regression. Newbury Park, CA: Sage Publications. Jaffe, E. D., & Ghymn, K. (2003). Importer vendor choice behavior: A comparative study. Journal of Global Marketing, 17(1), 5–22. Jones, G. R., & George, J. M. (1998). The experience and evolution of trust: Implications for cooperation and teamwork. Academy of Management Review, 23, 531–546. Kahn, W. A. (1990). Psychological conditions of personal engagement and disengagement at work. Academy of Management Journal, 33(4), 692–724. Katsikeas, C. (1998). Import stimuli and import development. Management International Review, 38(3), 215–237. Kenny, D. A., Kashy, D., & Bolger, N. (1988). Data analysis in social psychology. In: D. A. Kenny, D. Kashy & N. Bolger (Eds), Handbook of social psychology (pp. 233–265). New York: McGraw-Hill. Kim, K., & Frazier, G. L. (1997). On distributor commitment in industrial channels of distribution: A multicomponent approach. Psychology and Marketing, 14(8), 847–877. Kramer, R. M. (1999). Trust and distrust in organizations: Emerging perspectives, enduring questions. Annual Review of Psychology, 50, 569–598. Leonidou, L. C. (1989). The exporter–importer dyad – an investigation. Journal of Managerial Psychology, 4(2), 17–28. Liang, N., & Parkhe, A. (1997). Importer behavior: The neglected counterpart of international exchange. Journal of International Business Studies, 28(3), 495–530. Lye, A., & Hamilton, R. T. (2000). Search and performance in international exchange. European Journal of Marketing, 1(2), 176–189.

Antecedents of Relationship Commitment in an Import–Export Dyad

137

Marshall, R. S., & Boush, D. M. (2001). Dynamic decision-making: A cross-cultural comparison of US and Peruvian export managers. Journal of International Business Studies, 32(4), 873–893. Mason, C. H., & Perreault, W. D., Jr. (1991). Collinearity, power, and interpretation of multiple regression analysis. Journal of Marketing Research, 28, 268–280. Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of Management Review, 20, 709–734. McAllister, D. J. (1995). Affect- and cognition-based trust as foundations for interpretation. Academy of Management Journal, 38(1), 24–53. http://www.merriam-webster.com/ Morgan, R., & Hunt, S. (1994). The commitment–trust theory of relationship marketing. Journal of Marketing, 58(3), 20–38. Nijssen, E. J., Douglas, S. P., & Calis, G. (1999). Gathering and using information for the selection of trading partners. European Journal of Marketing, 33(2), 143–162. Nunnally, J. C. (1978). Psychometric theory. New York, NY: McGraw-Hill. Ragins, B. R., & Dutton, J. E. (2007). Positive relationships at work: An introduction and invitation. In: J. E. Dutton & B. R. Ragins (Eds), Exploring positive relationships at work: Building a theoretical and research foundation (pp. 1–25). Mahwah, NJ: Lawrence Erlbaum. Robinson, S. L. (1996). Trust and breach of the psychological contract. Administrative Science Quarterly, 41, 574–599. Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different after all: A cross-discipline view of trust. Academy of Management Review, 23, 393–404. Ryan, R. M., & Bernstein, J. H. (2004). Vitality. In: C. Peterson & M. E. P. Seligman (Eds), Character strengths and virtues: A handbook and classification (pp. 273–290). New York: Oxford University Press. Ryan, R. M., & Frederick, C. (1997). On energy, personality and health: Subjective vitality as a dynamic reflection of well-being. Journal of Personality, 65, 529–565. Simpson, J. A. (2007). Psychological foundations of trust. Current Directions in Psychological Science, 16(5), 264–268. Skarmeas, D., Katsikeas, C. S., & Schlegelmilch, B. (2002). Drivers of commitment and its impact on performance in cross-cultural importers–seller relationship: The importer’s perspective. Journal of International Business Studies, 33(4), 757–784. Spekman, R. E. (1988). Strategic exporter selection: Understanding long-term buyer relationships. Business Horizons, 31(4), 75–81. Styles, C., & Ambler, T. (2000). The impact of relational variables on export performance: An empirical investigation in Australia and the UK. Australian Journal of Management, 25(3), 61–281. The new titans: A survey of the world economy. (2006). The Economist, September 16. Zand, D. E. (1972). Trust and managerial problem solving. Administrative Science Quarterly, 17, 229–239.

This page intentionally left blank

CONSTRUCTING JURISDICTIONAL ADVANTAGE$ Maryann Feldman and Roger Martin ABSTRACT This chapter aims to advance economic development theory through the concept of jurisdictional advantage; demonstrating how places might strategically position themselves to gain economic advantage; then considering how this place-specific advantage might be constructed. We choose the term ‘‘jurisdiction’’ to define the set of actors that have a common interest in a spatially bound community. Jurisdictions are entities with a legitimate political ability to influence social and economic outcomes within their boundaries. Borrowing from the literature on corporate strategy, the uniqueness of local capabilities becomes a source of advantage for jurisdictions. We consider how to measure and construct jurisdictional advantage.

$

This chapter is a re-print of an article that was published in Feldman, M. P., & Martin, R. (2005). Constructing jurisdictional advantage. Research Policy, 34, 1235–1249, reprinted with permission from Elsevier.

New Perspectives in International Business Research Progress in International Business Research, Volume 3, 139–162 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1745-8862/doi:10.1016/S1745-8862(08)03007-0

139

140

MARYANN FELDMAN AND ROGER MARTIN

1. INTRODUCTION In the face of increasingly convincing evidence about the returns to agglomeration economies, and the strategic importance of location, pragmatic questions remain about how places may use this information to create economic growth. How can a given place in a particular situation with a limited set of resources pursue a strategy that will provide sustained economic results? Addressing this question requires prescriptive theory and conceptual frameworks. Moreover, perspectives for improving a region’s economic growth tend to focus on the role of government, underemphasizing the role of other actors that have significant local interests and play momentous roles. Actors such as business firms, arts and civic organizations, professional associations, universities and individual citizens are concerned about quality of life and want to influence the economic prospects of the communities where they live and work (Becherer, 2000). The relevant policy question is how to apply an appreciation of the benefits of location and cluster dynamics to implement strategies to promote economic growth. This question takes on greater urgency given structural changes in the world economy brought on by competitive pressures from newly emerging low cost counties. While corporate outsourcing allows firms to lower production costs technologically sophisticated firms compete on the basis of differentiated performance and innovation. And, it is in this case that certain locations confer an advantage that increases the productivity of investments made in innovative capacity by providing a platform to leverage key resources and relationships. Of course, this advantage is typically the result of specific unique characteristics that are built up over time to form a coherent place-specific activity set that is not easily transferred or replicated and forms the basis for sustainable advantage for both firms and industries (Feldman & Martin, 2005). This chapter aims to advance economic development theory on three fronts: introducing the concept of jurisdiction to the literature on geography and clusters; demonstrating how places acting as jurisdictions might strategically position themselves to gain economic advantage; and then considering how this place-specific jurisdictional advantage might be constructed. We choose the term ‘‘jurisdiction’’ to define the set of actors that have a common interest in a spatially bound community rather than using the more passive term of locational advantage. Jurisdictions, at a variety of spatial scales, are entities that have a legitimate political ability to influence outcomes within their borders. Section 3 of the chapter further defines our concept of jurisdiction, and considers the economic development

Constructing Jurisdictional Advantage

141

objective for which jurisdictions might realistically strive, with a discussion of measurable outcomes. Section 5 draws on corporate strategy literature to consider how communities may position themselves in order to create opportunities for economic growth. We argue that every jurisdiction has unique assets that may be built upon to construct a coherent activity set that would be difficult for others to copy. Considering examples of activity sets from Hollywood and the New York Fashion district, we examine the elements of the industry’s local advantage and how the industry became established. Section 6 considers how jurisdiction-specific advantage may be constructed and Section 7 concludes the chapter. The next section reviews the literature on the geography of innovation to provide a framework for considering jurisdictional advantage.

2. THE ENDOGENITY OF INDUSTRY–PLACE SUCCESS Economists, starting with Marshall in 1890, have long recognized the strategic importance of location to economic activity (Marshall, 1920). This is well-known rather than a historical artifact primarily because of the recent economic success of Silicon Valley and Route 128 (Saxenian, 1994; Kenney & von Burg, 1999). Even as the Internet provides real time connectivity, there is little dispute that global economic activity is agglomerating: industries are concentrating spatially in tightly defined geographic areas precisely because proximity and access matters. Empirical research increasingly provides evidence of the benefits that accrue to location (e.g. Aharonson, Baum, & Feldman, 2004; Almeida & Kogut, 1999; Baptista & Swann, 1998; Beaudry, 2001; Beaudry & Breschi, 2003; Glaeser, Kallal, Scheinkman, & Shleifer, 1992; Henderson, 1994; Sorenson, Rivkin, & Fleming, 2004; Rosenthal & Strange, 2003; Swann & Prevezer, 1996). As a result, a consensus is developing that compact geographic areas, typically centered in cities, are more important than countries or sub-national regions when considering economic growth and prosperity (Krugman, 1998; Fujita & Thisse, 1996). An important distinction is between the geographic concentrations of production and the location of innovation. Whereas the geographic concentrations of production is often due to the location of natural resources, ease of transportation or historical inertia, the location of innovation is due to knowledge externalities and subject to increasing returns (Audretsch & Feldman, 1996). While innovation yields greater

142

MARYANN FELDMAN AND ROGER MARTIN

productivity and the increases in wages that jurisdictions seek, jobs associated with routine production remain geographically in place as long as the physical investments are economically viable. Once physical assets are depreciated or obsolete, and if the market changes or costs become uncompetitive, these locations are easily abandoned. Being at the locus of activity for an industry promotes innovation. Certain locations supply localized knowledge externalities or spillovers that provide positive economic value but are beyond the ability of market mechanisms to price and efficiently allocate. The significance of localized knowledge spillovers as inputs to firms’ innovative activities suggest that their most creative and highest value-added activities do not proceed in isolation, but depend on access to new ideas. Location mitigates the inherent uncertainty of innovative activity: proximity enhances the ability of firms to exchange ideas and be cognizant of important incipient knowledge, hence reducing uncertainty for firms that work in new fields (Feldman, 1994). Innovation clusters spatially where knowledge externalities reduce the costs of scientific discovery and commercialization. In addition, firms producing innovations tend to be located in areas where there are necessary resources: resources that have accumulated due to a region’s past success with innovation. In this way, firm success and city economic growth are endogenous and mutually dependent. The cumulative nature of innovation manifests itself not just at firm and industry levels, but also at the geographic level, creating an advantage for firms locating in areas of concentrated innovative activity. These factors can generate positive feedback loops or virtuous cycles, as clusters attract additional specialized labor and other inputs, as well as the greater exchanges of ideas. What is critical is that these clustered industries tend to be more innovative and have greater productivity which is why we observe wage premium for such clusters (see for reviews Baptista, 1998; Audretsch & Feldman, 2005). The fact that clusters exist and provide economic benefit does not imply that they can be easily built – the costs associated with trying may outweigh realized benefits. The difficulty is trying to determine what activities may aid the formation and building of clusters and what activities waste or dissipate resources. While there is little disagreement over the proposition that clusters are economically advantageous, the notion that every place should attempt to build clusters is more controversial; proactive attempts to build clusters often fail or yield results that are different than anticipated (Feldman & Francis, 2004; Orsenigo, 2001). Typically, cluster-building initiatives focus on emerging, high growth industry with great political flourish and the substantial commitment of public resources. These efforts

Constructing Jurisdictional Advantage

143

are often mimicked by similar jurisdictions. For example, in the U.S. there are 40 states that have biotech initiatives and another nine states that list biotech as a target in their state technology plan. The only state missing of the 50 states in the U.S. is Indiana, but examination of the state budget reveals line items that focus on building biotech resources (Biotechnology Industry Organization, 2004, p. 28). However, the vast majority of cluster initiatives currently underway have little chance of achieving their articulated goals (Leslie & Kargon, 1994). In retrospect, the investment of time, energy and capital may not be justified. Of course, the threat of being left behind is great and localities have little choice but to participate in these races (MacRae, 2002). Incentives that dissipate the natural advantages that accrue to agglomerations may not only waste resources but may act to the detriment of innovation and technological change nationally or globally, as resources are pulled to places where they will be less productive and the benefits of external scale economies are lessened. An examination of the ways in which technologies and industry evolves reveal the importance of prior actions and the cumulativeness of advance or path dependencies (Arthur, 1990). The history of regions and industries become intertwined and it is difficult to transplant industries. Moreover, the path of emerging industries is difficult to predict and is extremely fluid. Our current understanding of an industry may not be able to anticipate future scientific developments, the temperament of consumers and their acceptance of a product and the directions that technology may take. When a technology reaches a stage when it can be easily understood and valued, the established centers may be described as first movers, already having an advantage over other locations. As a result, there is a tendency for activities which are ahead to get even further ahead. By the time an industry is well known enough to be targeted for economic development first mover jurisdictions have probably already captured the lion’s share of the benefits and are positioned for greater advantage, making it difficult to catch-up or overtake them following the same technology. Moreover, for clusters to be successful there is a need not to be captured by the prevailing logic of the industry but to adapt a new model that reflects a different vision for a technology or an industry.

3. DEFINING JURISDICTIONS When considering clusters, it is important to remember that the most cited example – Silicon Valley – is not a real place in any political, geological or

144

MARYANN FELDMAN AND ROGER MARTIN

administrative sense. Geographically, Silicon Valley defies standard classification – it is simply more that than the administrative units of Palo Alto or Menlo Park or even the larger unit of Santa Clara County. Silicon Valley, like other industry clusters, is a more amorphous concept that spans multiple administrative units. When we talk about clusters we are really talking about spatially defined epistemic communities of common interest. Most generally, we are talking about the construction of a common vision around an industry, technology or set of related interests. History is replete with examples of co-located firms defining technological frontiers (c.f. Allen, 1983; Maskell & Malmberg, 1999). The geographic concentrations of related entities create synergies that provide unique activity sets that promote the emergence of new industries: combining new knowledge with existing expertise is the essence of innovation. Sociologists such as Latour and Woolgar’s (1986) discussion of scientific discoveries or Abbate’s (1999) history of the Internet documents the cascading decisions that shape human creative work. This suggests that innovation is socially constructed and reflects a consensus vision of what is possible, a series of complementary activities and imitation and experimentation. Innovation is a creative activity influenced by the expectations, conversations and personalities of the individuals involved. Certainly, to the extent that social interactions are facilitated by co-location and frequent interaction, innovation becomes a local event. These epistemic communities self-organize and there are gains to frequent, face-to-face contact. While a specific address may connote prestige, it is proximity, access to concentrations of resources and venues for social relationships that matter most in promoting innovative activity, productivity growth and higher wages. Why then introduce the term jurisdiction? The term connotes a broader and more inclusive description of the economic assets and joint decisionmaking of a spatially defined system that contributes to economic prosperity. Most importantly, the concept of jurisdiction implies the development of laws, regulations, norms and conventions that provide systems of governance and innovation. A jurisdiction may be broad – for example, a country, in which case the vast majority of laws and regulations are the product of the jurisdiction. Or, a jurisdiction may also be narrow – for example, a city, in which case some of the laws and regulations are selfimposed while others are imposed by a super-ordinate body, such as the country. While the animate actors in the jurisdiction do not have complete

Constructing Jurisdictional Advantage

145

freedom of action, they certainly have a great deal of scope for decisionmaking and implementation. Jurisdictions include animate actors such as firms, governments/administrative institutions, educational institutions, cultural institutions and citizens as well as inanimate features such as natural endowments and physical infrastructure. Each of these animate and inanimate actors has the potential to contribute to the development – or lack thereof – of economically beneficial clusters. A physical asset, such as a great port, can contribute to the growth of a logistics cluster. Universities or government labs may or may not spin-off ideas that create new products, new firms and new industries (Bania, Eberts, & Fogarty, 1993). Great arts and cultural organizations can attract highly qualified personnel who value amenities and are critical human capital for certain industries (Florida & Gates, 2002). Norms supporting entrepreneurship can increase the number of start-up firms. Conversely, noncompete employment clauses or punitive bankruptcy laws may lessen entrepreneurial activity (Gilson, 1999; Becker & Hellmann, 2003). All of these are features that should be considered when analyzing a jurisdiction. What should be the goals of a jurisdiction? A basic metric is the prevailing wealth in the geographic area. Wealth is a combination of wages and investments. The greater the positive variance in wage levels from the national or worldwide mean, the greater the jurisdictional advantage from which residents benefit, other things being equal. Investment is an important measure of the wealth of the jurisdiction. For the majority of the population of the developed world, the single largest investment is home equity and the value of local amenities, quality of life and general economic outlook is capitalized in housing prices. Higher local wages and a growing local economy are expected to contribute to appreciation of real estate values and the wealth of property owners. Moreover, increase in property values yield higher tax revenues for the jurisdiction which, if used judiciously, increase amenities and public services. In this way, virtuous cycles of economic growth may be created. The logical step is to recognize that strategies to increase economic potential should focus on broadly defined jurisdictions and the construction of what we term jurisdictional advantage as a means to promote economic growth and prosperity. This chapter takes a next step toward a prescriptive theory of clustering and agglomeration by providing a theory of jurisdictional advantage. It takes these steps because the authors believe that the literature will not be able to create a useful prescriptive theory of jurisdictional advantage without these building blocks.

146

MARYANN FELDMAN AND ROGER MARTIN

4. WHAT IS A JURISDICTIONAL STRATEGY? The collective choices that actors make over time that shape the activity in a place define a jurisdictional strategy. This is not meant to imply that there is always clear and conscious coordination of the choices to produce a strategy. Sometimes the choices will be made with no coordination, the logical outcome of market forces. Nevertheless, the sum of these choices forms the basis for what the jurisdiction actually does. And as with corporations, strategy is choice; strategy is not public proclamations or planning processes, it is what an entity actually does. In this respect, every corporation has a strategy, whether that strategy is articulated or not. Similarly, every jurisdiction has a strategy which is defined by what its actors choose to do, whether in coordination or not and whether articulated or not. With this background, we define a successful jurisdictional strategy as one which produces relatively high and rising wages for the workers and real estate values for property owners within the boundaries of the jurisdiction. If wage levels are higher than other comparable jurisdictions, then the jurisdiction is translating its human, physical, and other capital into higher economic output per worker than its counterparts. And if wages are rising, then the jurisdiction is likely to be increasing its relative effectiveness rather than regressing toward the mean. Similarly, if real estate values are increasing, those already owning real estate in the jurisdiction are rewarded with appreciation of the value of their property. The longer a jurisdiction can maintain high and rising wages and values, the more valuable the jurisdictional strategy. We are particularly interested in jurisdictional strategy at the level of the city-region because the literature on clustering and agglomeration increasingly points to the importance of small and compact geographic units as being a critical element in the performance of industries. Countries and states provide institutional framework within which cities work and help define the opportunity set. However, differences in jurisdictional performance appear to derive from the choices made by local actors.

5. WHAT IS JURISDICTIONAL ADVANTAGE? The subsequent question is, what is the nature of a set of jurisdictional strategy choices that produces advantage evidenced by relatively high and rising wage and property value levels? Here we borrow from corporate strategy.

Constructing Jurisdictional Advantage

147

Over the past 30 years, the concepts of corporate strategy and strategic thinking have become well accepted by firms. The literature on corporate strategy finds that observed differences in firm performance are due to organizations capabilities which are unique and not easily imitated (Barney, 1991; Dosi, Nelson, & Winter, 2000; Nelson, 1991; Rumelt, 1984; Wernerfelt, 1984). While similar resources may be available to all firms, the ability to deploy these resources productively is not uniformly distributed and depends on the firms’ capabilities (Penrose, 1959). Moreover, these capabilities develop over time as the result of historically determined endogenous learning processes (Nelson & Winter, 1982). As a result of this path dependency, firm capabilities are unique and not easily replicated. The result is that specialized resources used by a firm are embedded in an organizational context and their effective use is contingent on other complementary assets (Rumelt, 1987). The mechanism through which a company produces advantage is an activity system which is unique, not easily replicated and sustainable.

5.1. Activity Systems: Borrowing from Corporate Strategy One school of corporate strategy holds that advantage results either from lower cost relative to the firm’s competition or the production of a set of attributes that are uniquely valued by the market (Porter, 1980, 1985). These advantages are based on the construction of unique activity systems, which are defined as a coherent web of activities. Taken together, these activity systems provide an advantage because the individual activities and components fit well together and actually reinforce each other. The fit and reinforcement makes it difficult, if not impossible, for competitors to replicate a successful firm’s strategy – a rival firm would have to replicate every single aspect of the activity system in order to match the advantaged firm’s strategy. Thus, the essence of strategy is to construct an activity system that allows the firm to perform differently than the competition or to perform different activities than the competition. Consider the example of Southwest Airlines, which has been the most successful airline in the U.S. market over the past 30 years, in terms of the level of profitability, stability of earnings and growth in market share (Porter, 1996). Southwest is noted for flying a completely standardized fleet of Boeing 737s, flying to and from secondary airports, having the most frequent number of daily departures from each of its locations, and utilizing the Internet rather than travel agents for booking. Importantly, its

148

MARYANN FELDMAN AND ROGER MARTIN

Fig. 1.

Southwest Airlines: low cost advantage. Source: Porter (1996).

exemplary performance outcomes are not the result of any single thing that Southwest does. As demonstrated in Fig. 1, Southwest achieves advantage by performing a set of activities in a comprehensive manner that fit together and reinforce each other to produce a significant and sustainable cost advantage over its competitors. To challenge Southwest Airline’s success, a potential competitor would have to match every single aspect of Southwest’s activity system. An activity system can provide a low cost advantage by enabling the firm to produce a product or service that is roughly equivalent to the competition but at a significantly lower cost. This results in higher profitability than the average competitor and has been the case for Southwest in the airline industry. Any firm with a cost advantage is able to set prices in the industry and ultimately force marginal firms who do not have a similar cost advantage out of the industry. It is important to note however, that being a low cost firm is not the same as being the firm offering the lowest price. Having the same cost structure as competitors and deciding to sell at a lower profit margin is not a strategy for long-term advantage. Indeed, it is a recipe for transferring economic value from corporate shareholders to customers. This is simply not a sustainable

Constructing Jurisdictional Advantage

149

long-term strategy. Any competitive firm in the same industry can simply cut its own prices and margins to compete in the short run. This, ultimately, leads to a race to the bottom in terms of profitability as firms successively continue to lower prices and squeeze profit margins. The second strategy for corporate advantage is product differentiation – producing a product or service that is considered to be uniquely valued by a segment of customers and for which those customers are willing to pay a premium price. For example, Progressive Insurance, as demonstrated in Fig. 2, offers a differentiated automobile insurance service to a nonstandard segment of drivers (Porter, 1996). It offers quotes that are bettertailored to the true risks associated with different categories of drivers and provides quick and easy settlement of claims using an extensive fleet of vanbased adjusters. Like Southwest Airlines, Progressive also has a distinctive activity system that features many reinforcing activities, such as its sophisticated pricing database, a fleet of claims-settling vans, and unique training and compensation structures, which fit together to produce a service that is highly valued by its customers and is produced at a competitive cost.

Fig. 2.

Progressive insurance: differentiation advantage. Source: Derived from Scott (2004).

150

MARYANN FELDMAN AND ROGER MARTIN

An effective differentiation strategy also requires operational efficiency and a competitive cost structure. Firms that charge a premium price but do not maintain an efficient cost structure end up with lower profit margins and do not have the resources to reinvest in sustaining long-term advantage. An effective differentiation strategy requires an activity system that is efficient in terms of cost structure while simultaneously producing a differentiated product or service. Strategy allows firms to define their objectives and make decisions that focus their energies and investments. Most importantly, strategy allows firms to know what activities are not going to realistically advance their goals.

5.2. Jurisdictional Activity Sets Following this logic, it seems that places might also have similar capabilities that develop over time, are unique and not easily replicated. Jurisdictions may benefit from adapting a similar strategic orientation and building an activity system that is unique and valuable in order to produce either a low cost or differentiation advantage over other jurisdictions. Next, we look at Hollywood and the New York garment industry as two examples of how jurisdictions have built such activity systems. Hollywood is so well known that its name connotes both an industry and a geographic place. Hollywood has been the global center for film production since the 1920s and has maintained its dominance through changes in production technology and business models (Scott, 2005). In 2001, the southern California entertainment cluster, as defined by The Institute for Strategy and Competitiveness, employed 178,000 people, or 16% of national employment for this industry. The average wages in the entertainment industry are US $60,000 which was 50% higher than the national average wage for the industry and greater than three times the average wage for the region. Scott (2004, 2005) rejects the conventional explanation that the industry located in southern California was due to favorable weather conditions and scenic vistas, but instead offers a view that suggests a coherent activity set created the industry’s dominance. At the turn of the century, the then dominant New York-based Motion Picture Patent Trust priced its films by the foot, a policy that gave producers little incentive to raise quality or innovate (Scott, 2005, p. 17). In contrast, Hollywood independent producers concentrated on distinctive feature films and promoted individual stars and the development of methodological procedures to break down the shooting

Constructing Jurisdictional Advantage

151

process into disconnected segments that were reassembled later. This lowered costs and gave rise to the studio system with its advanced division of labor and sophisticated managerial model of production. Hollywood was able to differentiate itself from New York with an activity set that included new business models, a cost advantage and the ability to experiment and innovate. Fig. 3 diagrams the activity set that established Hollywood’s advantage, ca. 1928, or about the time when Koszarski (1990) notes that the term ‘‘Hollywood’’ was used in a generic sense to refer to the motion picture industry. A supporting infrastructure developed that reinforced the activity set with skilled crafts such as film editing, file laboratories, agents, orchestras, costumes and props. In addition, the open shop labor market arrangements in Los Angeles, which were in place until 1935, provided flexibility. The Academy of Motion Pictures Arts and Science was created in 1927 as an overarching union with five branches representing producers, writers, directors, actors and technicians. The Academy, well known for the Oscar Awards, encourages artistic achievement and diffused best practices.

Fig. 3.

New York City women’s garment production. Source: Derived from Rantisi (2002a, 2002b).

152

MARYANN FELDMAN AND ROGER MARTIN

Another example is the New York City fashion industry, one that has become synonymous with style (Rantisi, 2002a, 2002b). The New York apparel industry had a modest beginning on the lower east side of Manhattan, the site of skilled immigrant communities in the 1880s. Most large cities had garment districts at the time and there was nothing particularly unique about New York. By 1920 the industry migrated to the Garment District in the western part of midtown where it is currently located. In these intervening years, the focus shifted to high value-added fashion (Rantisi, 2002b, pp. 447–448). Currently, the New York cluster accounts for 40% of U.S. value added for the category of women and children’s fashions (Rantisi, 2002b, p. 442). Although Los Angeles has greater employment in this industry category, wages in New York are more than double the national average for the industry. Fig. 3 provides the activity set that reinforces the prominence of the fashion industry in New York. The development of a range of retailing formats in New York – from department stores like Macy’s, Bloomingdales and Lord and Taylor’s, as well as specialized boutiques such as Henri Bendel – helped to develop varied markets for the industry’s output. Fashion magazines, notably Women’s Wear Daily (now WWD), Harpers Bazaar and Vogue helped establish a fashion culture and disseminate industry trends to the national market. This was supported by the developing advertising industry and expertise in fashion photography. The international Ladies Garment Workers Union helped promote safe working conditions and standardized wages. This ensured high-quality goods, encouraged women to upgrade their skills and also enabled workers to become active consumers (Rantisi, 2002b, p. 448). Supporting educational institutions such as the Pratt Institute, the Parson School of Design and the Fashion Institute of Technology ensured a steady supply of skilled labor. In addition, the production system came to be characterized by a large number of specialized contractors, job shops and specialty fabric designers. These small, niche firms provide a flexibility and culture of experimentation. In conclusion, these are two examples of jurisdictional advantage. Similarly, the Boston Biotech Cluster (Owen-Smith & Powell, 2007) or Silicon Valley (Kenney & Patton, 2005) also have a coherent activity set that connect biotech firms to venture capitalists, specialized business services (such as contract research organizations), legal services and universities. However, these examples illustrate the importance of local labor agreements, experimentation with new business models, the development of a local division of innovative labor and other institutional linkages which are specific and idiosyncratic. Taken together, these activities comprise the local

Constructing Jurisdictional Advantage

153

activity set that provides an economic advantage to companies, industries and jurisdictions. Moreover, these examples demonstrate the path dependencies and cumulativeness of the activity sets (Arthur, 1996). Innovation, the ability to introduce novel ideas and consistently improve performance, is a notable characteristic of these locations and a key to their economic performance. In the examples examined above, jurisdictions were able to translate an initial cost advantage into a sustained competitive advantage through the construction of an activity set that yielded an advantage based on differentiation. This is to say, the interaction of the features of the activity set, make Hollywood (in film) and New York (in fashion) jurisdictions that provide unique value to a firm in the respective industry, such that locating in that jurisdiction is uniquely valuable vis-a`-vis alternative jurisdictions. Moreover, the two industries examined here are illustrative as they are creative industries where human capital known as talent is important (Caves, 2002). In a global economy where unskilled labor is inexpensive, transportation and communication costs are negligible and raw material matter little, talent is becoming an important competitive asset (Florida & Gates, 2002). The current obsession with outsourcing gives rise to the question of whether low wage regions enjoy jurisdictional advantage. We argue that they do not because the manifestation of jurisdictional advantage is high and rising wages. A jurisdictional strategy based on low wages would undermine itself on the way to jurisdictional advantage. That is, in order to move toward advantage, wages would have to rise, unraveling the low wage approach. A low wage policy in the realm of jurisdictional strategy is analogous to a low price approach in corporate strategy. The targeted beneficiaries of corporate strategy – the shareholders – do not benefit from a low price strategy; the customers do. The targeted beneficiaries of jurisdictional strategy – workers in the jurisdiction – do not benefit from a low wage strategy. In particular, decreasing wages is not a strategy for long-term competitive advantage in industries where talent and human capital are important. Producing talent requires investments in human capital and incentives that motivate individuals to create and engage their talents. Moreover, talented individuals do not work alone. Indeed, talent may only be recognized in a specialized setting where genius may be appreciated and appropriately rewarded. Hollywood demonstrates that innovation in business models that accompanied technological breakthroughs in moviemaking were critical – a vision of what the motion picture industry might be. While certain

154

MARYANN FELDMAN AND ROGER MARTIN

individuals stand out as influential, the activity set reflects the efforts of many people, acting individually and collectively. The New York fashion industry demonstrates the success of building on rather pedestrian activities. Interestingly, both industries made important transitions in the form of advantage they enjoyed. Each began with a low cost advantage – it was less expensive for firms to do business in those industries in those jurisdictions. This was not because of low wages, but rather because of other cost reducing activities. However, both industries eventually developed differentiation advantages – with activity systems that provided an environment that firms in the industry could not find elsewhere. What allowed these jurisdictions to differentiate themselves is the construction of reinforcing activity sets that both promote and further define the industry. Of course, this advantage is typically the result of specific unique characteristics that are built up over time to form a coherent place-specific activity set that is not easily transferred or replicated and forms the basis for sustainable advantage for both firms and industries (Feldman & Martin, 2005). Without engaging a variety of different agents and supporting institutions these jurisdictions may not have achieved primacy within their industries. The examples presented here intended to be illustrative. The tendency might be to dismiss these examples as offering unrealistic aspirations for average locations. Yet the literature contains a large number of detailed and carefully constructed case studies about different industries and the genesis of the jurisdictions in which they reside. The heuristic of activity systems presented here and the study of coherent activity systems may be used to consider elements that are missing in certain locations that are less successful. Moreover, rather than simply replicating existing successful clusters when we consider these activity systems, it becomes apparent that activities are reinforcing and that the success of the industry and jurisdiction co-evolve. This chapter will now sketch out some ideas on how the animate actors in a jurisdiction can use the theory of jurisdictional advantage to construct jurisdictional advantage.

6. CONSTRUCTING JURISDICTIONAL ADVANTAGE Constructing jurisdictional advantage is not easy, just as it is not simple for companies to craft strategies that yield long run growth and profitability. However, there are a few things that can be noted. The evidence that firm location decisions are not responsive to jurisdictional tax differentials

Constructing Jurisdictional Advantage

155

except at the intrametropolitan area (Bartik, 1991; Papke, 1991) suggests that individual municipalities may gain if they drop their tax rates or offer special incentives. Of course, this creates artificial competition between administrative units that may be part of the same industry cluster and may be viewed as a unified jurisdiction in terms of common economic objectives. These actions, while yielding short-term benefits, may result in a race to the bottom. Individual municipalities may achieve long-term benefit if they view themselves as subsidiaries or divisions of a larger geography and cooperate to their mutual advantage. Jurisdictional activity system is not the product of any one class of actors – not firms, not individuals, not governments, not universities. Highly competitive clusters and therefore highly advantaged jurisdictions have activity systems in which multiple actors play interacting and reinforcing roles to produce and reinforce its uniqueness. Each of the actors pursues their own self-interests. People choose to work where they can maximize their own wealth and happiness. Firms invest in things that have the prospect of increasing their returns. Arts and cultural organizations pursue self-expression. Universities follow the desires of their scholars. Governments do the things they need to do to get re-elected. However, each of these self-interested needs is furthered by a jurisdictional activity system that produces high and rising wages and real estate values, so each actor is working toward their self-interest when it contributes to jurisdictional advantage. However, it has to be recognized that the actions of any one actor will only go so far. Individual firms, for example, will not invest meaningfully in things that payoff primarily for other firms or actors in the system. Governments need to fund those items for which there are high externalities from the perspective of any individual firm. Because no one actor can do everything and each actor has an interest in experiencing jurisdictional advantage, constructing jurisdictional advantage takes the will of all the actors – a consensus vision and vision of uniqueness. Each party needs to ask what unique contribution it can make to the unique activity system. Governments are in the best position to play a coordinating role. Two extreme philosophies anchor the ends of the jurisdictional policy spectrum. At one end of the spectrum lies aggressive centralized planning that targets an industry or industries and puts in place a strategy to attract that industry or industries to the jurisdiction, often with tax incentives. There are myriad examples where politicians and civic leaders focus on some emerging, high-growth industry with greater fanfare. This is certainly the case with current example of the 49 states with biotech industry initiatives.

156

MARYANN FELDMAN AND ROGER MARTIN

Even when efforts are successful at generating start-up companies it is difficult for a jurisdiction to garner long-term benefits if complementary assets are lacking (Connecticut Center for a New Economy, 2004). The opposite extreme is to let market forces determine the allocation of resources, a straightforward laissez faire philosophy. The underlying rationale of this philosophy is that industrial clusters that are part of successful cities arise for a variety of historically contingent or serendipitous factors that are not easily replicated. Firms locate and invest in particular cities for reasons that are not well understood, much less predictable and controllable. This view suggests that the most constructive thing a jurisdiction can do is let market forces determine its future. However, just as there are logical problems with the centralized, command and control end of the spectrum, there are weaknesses to the laissez faire argument. First, it is difficult to find examples of highly successful jurisdictions or even individual clusters in such jurisdictions where there was little evidence of involvement by actors other than firms – in particular involvement by governments acting in coordination with the provision of externality-laden investments. Even Silicon Valley, often used as the paragon of a jurisdiction driven by free market forces, would not have developed as a highly advantaged jurisdiction with among the highest wage levels and real estate values in the world without massive investment by governments in higher-education and research focused on the industries that have made Silicon Valley the envy of virtually all jurisdictions. Indeed Silicon Valley exemplifies the challenges in the increasingly knowledge-based economy where market failures lead to under-allocation of the very goods that provide advantage without coordinative activities. After all, this is one of the classic reasons for government provision of infrastructure, funding of basic research and promotion of public goods such as education. These resources, which are associated with market failure, take on new importance in the emerging knowledge-based economy and suggest that there is a role for collective action and government participation. The idea that firms act alone as solo players is a romantic image that has no good exemplars. Laissez faire proponents tend to point to high tech industries like software because they feature cutthroat competition and rapid entry and exit. However, it is one of the industries most dependent on and linked to the U.S. higher education system. This holds true for pharmaceuticals, medical services, aerospace and virtually all high-wage industries. Strategy is choice and closer inspection reveals that each of these industries has benefited from a jurisdictional activity set that

Constructing Jurisdictional Advantage

157

coordinates the choices of various categories of actors. Given that this is the case, having a goal – high and rising wages – and a structure for thinking about achieving that goal – differentiation or low cost – and a tool for guiding the strategy – a distinctive activity system – can be useful. Strategy has utility in suggesting that a route between the two ends of the spectrum is likely to be most effective. It suggests that a strict laissez faire approach is unlikely to provide the externality-laden assets necessary for the jurisdiction to succeed. However, it also restricts the overly expansive government coordinators. Government intervention can only be justified if it contributes to the development of a jurisdictional activity system that provides differentiation or low cost advantage to the jurisdiction in attracting, retaining and growing firms. Rigorous analysis of the government activities in cluster initiatives would undoubtedly indicate that the vast majority of them do not even have as a realistic goal, let alone a stated goal, of creating such an outcome. As we know from corporate strategy, investments made by businesses to simply replicate what other firms have already done or improve a firm’s competitive position from significantly behind to merely lagging are typically investments that do not earn an acceptable return on capital. Similarly, we would not expect government investments that are not made to support an advantaged jurisdictional strategy to have a positive net present value. As such the framework of jurisdictional advantage provide guidance for the difficult challenge of choosing coordinative and costly interventions that have a sufficiently high probability of paying off. What is the role of firms in jurisdictional advantage? A firm can act simply as a taker and exploiter of a jurisdiction. However, a firm is better served by being an active partner in jurisdictional advantage rather than a passive taker. The existence of externalities critical to firm success suggests that firms can receive benefits from the jurisdictional activity system that are outside of the market mechanism to price. As soon as it has made investments in a jurisdiction, a firm has an incentive to encourage other actors in the jurisdiction to build externality-laden investments so that the jurisdiction provides more advantages for itself in the future. While it may be argued that firms pay more taxes as a result of the higher profits they earn as a result of externalities, a higher level (not rate) of taxes is a small price to pay for increased competitive success and sustainability. In fact, it may be argued that firms may actively cultivate the sources of the agglomerative benefit by investing in local universities and cultural organizations, building infrastructure and so on. Moreover, these investments are typically tax deductible and provide a means to make targeted

158

MARYANN FELDMAN AND ROGER MARTIN

investments in jurisdictions rather than relying on the process of government budgeting. That is to say, those firms may actively build the external resources and infrastructure that benefit their bottom line, but all in the context of strengthening the jurisdictional activity system. Industrial recruitment incentives with special tax breaks and various other inducements that lower the costs of doing business are not low cost strategies. The evidence is that this type of strategy is a race to the bottom in a zero-sum game (Bartik, 1991). There is no evidence that it eventually leads to higher wages, which is the measure of a successful low cost jurisdictional strategy. In many respects, governments that use tax breaks or other monetary inducements are identical to firms that use sales as central tenants of their strategies. Sales are nothing but periodic low price strategies and low price without low cost is not a sustainable strategy. This is a difficult lesson that Sears and Kmart, who have now merged in an attempt to survive Walmart’s relentless market penetration, have learned. Sears and Kmart have historically followed what are called ‘‘hi-lo’’ merchandising approaches. That is, they have a normal price for their products, which is set at a sufficiently high level to earn a reasonable profit if they sell a targeted amount of products at that price. However, to drive volume in the stores, they hold sales by which they discount some or all merchandise for periods of time. Unfortunately for them, customers figure out this approach and disproportionately wait for goods to go on sale which simply reduces the overall margin on the portfolio of goods to a level that does not allow sufficient investment in upgrading by Sears and Kmart. Walmart entered the fray against Sears and Kmart with a different merchandizing strategy, which focused on cost structure, through leaner overheads, better distribution systems and narrower product selection. This was known as EDLP – or every day low pricing – merchandising strategy by which its everyday low price met or beat Sears’ and Kmart’s sale prices. As a result consumers could shop at Walmart every day and be confident they were buying the goods in their basket at a price that was below Sears and Kmart, even if Sears or Kmart had the product on sale. With their higher cost structures, Sears and Kmart could not compete and were battered in the process: Kmart declared Chapter 11. Jurisdictions that offer tax incentives or cash subsidies to attract new firms to the jurisdiction are following what may be called a modified hi-lo strategy. It is modified in that rather than offering alternative ‘‘hi’’ and ‘‘low’’ pricing to all customers at rotating times, the jurisdiction offers at

Constructing Jurisdictional Advantage

159

the same time a higher price scenario to some ‘‘customers’’ – i.e. firms that are already in the jurisdiction – and a lower price scenario to other ‘‘customers’’ – i.e. firms that it is trying to attract to the jurisdiction. As with Sears and Kmart, for the gambit to work, the jurisdiction desperately needs a large volume of high price sales in order to fund the discounts or tax breaks. This means that existing firms must to be taxed at a higher rate to fund the ‘‘low’’ pricing – in effect a subsidy from the current firms in the jurisdiction. As with Sears and Kmart, the ‘‘hi’’ sales are extremely vulnerable when they realize that they are paying higher prices. This both hurts the competitiveness of the existing firms and tends to drive them away, perhaps to another jurisdiction that will offer them lower prices. As with Sears and Kmart, ‘‘hi-lo’’ is not a robust strategy but a dangerous act of desperation that begs for a crushing competitive response.

7. REFLECTIVE CONCLUSIONS The pursuit of jurisdictional advantage is not without its challenges because there are so many factors that influence the outcomes. Given that future prosperity and quality of life in local communities are at stake, the questions of how this might be done are of more than academic interest. Previous work on clusters has emphasized the random nature of geographical location (Krugman, 1991; Klepper, 2004) suggesting that clusters arise from serendipitous events. In contrast, we argue that clusters may be constructed, but not in the way that policy typically proceeds by targeting an industry that is poised to take off in another location. Instead we argue that policy may be fruitfully employed by building upon unique placespecific assets. To develop the concept and practice of jurisdictional advantage, more work needs to be done on the success models of building jurisdictional advantage through this middle path between laissez faire and command and control. In particular, work needs to be done on the way in which governments can constrain themselves to activities that focus on providing the externality-laden investments without discouraging firms from investing where they are most capable of investing. Also, future work needs to determine how governments, businesses and other actors may work together to coordinate jurisdictional strategy. These are the next pieces of work to be done and we hope that this effort will motivate others to join in this pursuit.

160

MARYANN FELDMAN AND ROGER MARTIN

REFERENCES Abbate, J. (1999). Inventing the Internet. Cambridge, MA: The MIT Press. Aharonson, B. S., Baum, J. A. C., & Feldman, M. P. (2004). Industrial clustering and the returns to inventive activity: Canadian biotechnology firms, 1991–2000. Danish Research Unit on Industrial Dynamics (DRUID). Working Paper 2004-3. ISBN: 87-7873-153-4. Allen, R. C. (1983). Collective invention. Journal of Economic Behavior and Organization, 1(4), 1–24. Almeida, P., & Kogut, B. (1999). Localization of knowledge and the mobility of engineers in regional networks. Management Science, 45, 905–917. Arthur, B. (1996). Increasing returns and the new world of business. Harvard Business Review, 74, 100–110. Arthur, W. B. (1990). Silicon Valley locational clusters: When do increasing return imply monopoly? Mathematical Social Sciences, 19, 235–251. Audretsch, D. B., & Feldman, M. P. (1996). R&D spillovers and the geography of innovation and production. American Economic Review, 86, 630–640. Audretsch, D. B., & Feldman, M. P. (2005). The geography of innovation and spillovers. In: V. Henderson & J. F. Thisse (Eds), Handbook of regional and urban economics: Cities and geography (Vol. 4, pp. 2713–2739). North-Holland: Elsevier Science Publisher. Bania, N., Eberts, R. W., & Fogarty, M. S. (1993). Universities and the startup of new companies: Can we generalize from Route 128 and Silicon Valley? Review of Economics and Statistics, 75(4), 761–766. Baptista, R. (1998). Clusters, innovation and growth: A survey of the literature. In: G. M. P. Swann (Ed.), The dynamics of industrial clusters: International comparisons in computing and biotechnology (pp. 13–51). Oxford: Oxford University Press. Baptista, R., & Swann, P. (1998). Do firms in clusters innovate more? Research Policy, 27, 525–540. Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17(1), 99–120. Bartik, T. J. (1991). Who benefits from state and local economic development policies? Kalamazoo, MI: Upjohn Institute for Employment Research. Beaudry, C. (2001). Entry, growth and patenting in industrial clusters: A study of the aerospace industry in the UK. International Journal of the Economics of Business, 8, 405–436. Beaudry, C., & Breschi, S. (2003). Are firms in clusters really more innovative? Economics of Innovation and New Technology, 12, 325–342. Becherer, H. W. (2000). A more cosmopolitan way of life: Why local economic development matters. Vital Speeches of the Day, 66(15), 473. Becker, R., & Hellmann, T. (2003). The genesis of venture capital: Lessons from the German experience. CESIFO Working Paper No. 883. Biotechnology Industry Organization. (2004). Laboratories of innovation: State bioscience initiatives 2004. Washington, DC: Biotechnology Industry Organization. Caves, R. (2002). Creative industries: Contracts between art and commerce. Chicago: University of Chicago Press. Connecticut Center for a New Economy. (2004). Incubating Biotech: Yale Prospers, New Haven Waits, March. Dosi, G., Nelson, R. R., & Winter, S. G. (2000). Introduction. In: G. Dosi (Ed.), The nature and dynamics of organizational capabilities (pp. 1–22). New York: Oxford University Press.

Constructing Jurisdictional Advantage

161

Feldman, M. P. (1994). The geography of innovation. Boston: Kluwer Academic Publishers. Feldman, M. P., & Francis, J. (2004). Homegrown solutions: Fostering cluster formation. Economic Development Quarterly, 18, 127–137. Feldman, M. P., & Martin, R. (2005). Jurisdictional advantage. In: A. B. Jaffe et al. (Eds), Innovation policy and the economy (Vol. 5, pp. 57–85). Cambridge: MIT Press. Florida, R., & Gates, G. (2002). Technology and tolerance. The Brookings Review, 20, 32–64. Fujita, M., & Thisse, J.-F. (1996). Economics of agglomeration. Proceedings of the 10 Japanese International Economy, 339, 341. Gilson, R. J. (1999). The legal infrastructure of high technology industrial districts: Silicon Valley, Route 138, and covenants not to compete. New York University Law Review, 74(3), 575–629. Glaeser, E. L., Kallal, H. D., Scheinkman, J. A., & Shleifer, A. (1992). Growth in cities. The Journal of Political Economy, 100(6), 1126–1152. Henderson, J. V. (1994). Externalities and industrial development. Cityscape: A Journal of Policy Development and Research, 1(1), 75–93. Kenney, M., & Patton, D. (2005). Entrepreneurial geographies, support networks in the three high technology industries. Economic Geographies, 81, 201–228. Kenney, M., & von Burg, U. (1999). Technology, entrepreneurship and path dependence: Industrial clustering in Silicon Valley and Route 128. Industrial and Corporate Change, 8, 67–103. Klepper, S. (2004). Agglomeration through Spinoffs: How Detroit became the capital of the US automobile industry. Working Paper. Carnegie Mellon University. Koszarski, R. (1990). An evening’s entertainment: The age of the silent feature picture (pp. 1915– 1928). New York: Scribner’s. Krugman, P. (1991). Geography and trade. Cambridge, MA: MIT Press. Krugman, P. (1998). Space: The final frontier. Journal of Economic Perspectives, 12(2), 161–174. Latour, B., & Woolgar, S. (1986). Laboratory life: The construction of scientific facts. Princeton, NJ: Princeton University Press. Leslie, S., & Kargon, R. (1994). Electronics and the geography of innovation in post-war America. History and Technology, 11, 217–231. MacRae, M. (2002). Biotech growth: Will your city be a leader or a loser? Pharmaceutical Technology, 106, 106. Marshall, A. (1920). Principles of economics. London: Macmillan. Maskell, P., & Malmberg, A. (1999). Localised learning and industrial competitiveness. Cambridge Journal of Economics, 23(2), 167–185. Nelson, R. R. (1991). Why do firms differ, and how does it matter? Strategic Management Journal, 12, 61–74. Nelson, R. R., & Winter, S. (1982). An evolutionary theory of economic change. Cambridge, MA: Belknap Press. Orsenigo, L. (2001). The (failed) development of a biotechnology cluster: The case of Lombardy. Small Business Economics, 17(1/2), 77–92. Owen-Smith, J., & Powell, W. W. (2007). Accounting for emergence and novelty in Boston and Bay Area biotechnology. In: P. Braunerhjelm & M. P. Feldman (Eds), Cluster genesis: The origins and emergence of technology-based economic development (pp. 61–83). Oxford: Oxford University Press. Papke, L. E. (1991). Interstate business tax differentials and new firm location: Evidence from panel data. Journal of Public Economics, 45, 47–68.

162

MARYANN FELDMAN AND ROGER MARTIN

Penrose, E. T. (1959). The theory of the growth of the firm. New York: Wiley. Porter, M. E. (1980). Competitive strategy: Techniques for analyzing industries and competitors. New York: Free Press. Porter, M. E. (1985). Competitive strategy: Creating and sustaining superior performance. New York: Free Press. Porter, M. E. (1996). What is strategy? Harvard Business Review, 74(6), 61–78. Rantisi, N. M. (2002a). The local innovation system as a source of variety: Openness and adaptability in New York City’s garment district. Regional Studies, 36, 587–602. Rantisi, N. M. (2002b). The competitive foundations of learning and innovation: The case of women’s garment production in New York City’s garment district. Economic Geography, 78, 441–462. Rosenthal, S. S., & Strange, W. C. (2003). Geography, industrial organization, and agglomeration. The Review of Economics and Statistics, 85(2), 377–393. Rumelt, R. P. (1984). Toward a strategic theory of the firm. In: R. B. Lamb (Ed.), Competitive strategic management (pp. 556–570). Englewood Cliffs, NJ: Prentice-Hall. Rumelt, R. P. (1987). Theory, strategy, and entrepreneurship. In: D. Teece (Ed.), The competitive challenge (pp. 137–158). Cambridge, MA: Ballinger. Saxenian, A. (1994). Regional advantage. Cambridge, MA: Harvard University Press. Scott, A. J. (2004). Origins and growth of the Hollywood motion picture industry: The first three decades. Paper prepared for the conference on Cluster Formation and Development, Waxholm, Sweden, July 1–3, 2004. Scott, A. J. (2005). On Hollywood: The place, the industry. Princeton, NJ: Princeton University Press. Sorenson, O., Rivkin, J. W., & Fleming, L. (2004). Complexity, networks and knowledge flow. Working Paper. Anderson School of Management, UCLA and Harvard Business School. Swann, P., & Prevezer, M. (1996). A comparison of the dynamics of industrial clustering in computing and biotechnology. Research Policy, 25, 1139–1157. Wernerfelt, B. (1984). The resource-based view of the firm. Strategic Management Journal, 5, 171–180.

PART III THE GENESIS OF INTERNATIONAL ENTREPRENEURS

This page intentionally left blank

COMPARATIVE INTERNATIONAL ENTREPRENEURSHIP: THE SOFTWARE INDUSTRY IN THE INDIAN SUB-CONTINENT Shameen Prashantham, Amer Qureshi and Stephen Young ABSTRACT Purpose – In this chapter, we seek to extend understanding of the ‘international’ dimension of comparative international entrepreneurship (IE), by undertaking exploratory empirical research within a global industry viz. the software industry, and focusing on two local ecologies, namely a regional agglomeration (Bangalore, India) and less developed niche (Lahore, Pakistan) about which little is known. Methodology – On the basis of in-depth interviews in Bangalore and Lahore, exemplar case studies from both sub-national regions are presented, which highlight the relative significance of local milieu and ethnic ties in IE. Findings – The global nature of the software industry and the central role of the innovative milieu in the USA have important implications for the comparative IE literature. These refer particularly to the coordination

New Perspectives in International Business Research Progress in International Business Research, Volume 3, 165–187 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1745-8862/doi:10.1016/S1745-8862(08)03008-2

165

166

SHAMEEN PRASHANTHAM ET AL.

and integration of the entrepreneurial processes of opportunity discovery, evaluation and exploitation across frontiers. Close inter-milieu links provide the opportunity to access complementary assets and business networks. While Indians have influenced the development of Silicon Valley, their ties with Bangalore seem primarily to be based on hardnosed business relations. But in relation to Pakistan, while the US milieu is critical for all aspects of the entrepreneurial process, closed networks may be a barrier to long-term growth. Originality/value of chapter – Where our study goes beyond the literature is by highlighting the role of cross-border linkages between milieux.

INTRODUCTION Although large multinational enterprises are often deemed to be the major actors in the globalized economy, there is recognition that entrepreneurial younger firms play a significant role, particularly in global industries such as information technology (Young, Dimitratos, & Dana, 2003). Indeed, following Oviatt and McDougall’s (1994) conceptualization of international new ventures, considerable research interest in the field of international entrepreneurship (IE) has been generated (Oviatt & McDougall, 2005). Consistent with the views of Buckley and Ghauri (2004) on international business research in general, Zahra (2005) has pointed out that a deficiency in the IE literature is the neglect of the influences of the institutional environment and economic geography. A helpful framework through which to address this deficiency has been developed by Baker, Gedajlovic, and Lubatkin (2005) who address the topic of comparative IE. Baker et al. (2005) argue that the comparative social and institutional context in which new ventures are developed profoundly influence the manner in which they discover, evaluate and exploit new opportunities. In particular, they identify the role of the sub-national local context of the venture as being vitally important. They distinguish among three types of local context: regional agglomerations (mostly associated with advanced economies), broadly developed niches and less developed niches (mostly associated with developing economies). In this chapter we seek (a) to extend Baker et al.’s (2005) conceptual framework, emphasizing particularly the ‘international’ dimension of comparative IE, (b) by undertaking exploratory empirical research within

Comparative International Entrepreneurship

167

a global industry and (c) focusing on two local ecologies, namely a regional agglomeration and a less developed niche about which little is known. The research objective of the chapter is thus to explore the notion of comparative IE using the software sector in India and Pakistan as an illustration. We focus specifically upon the software industries in Bangalore and Lahore, which are arguably the most important software locations in the two countries; and represent a regional agglomeration (at least by emerging economy standards) (Baker et al., 2005) and a less developed niche, respectively. These sub-national regions constitute an interesting setting given that they share a common colonial past but yet have had contrasting institutional environments over the past six decades since 1947 when the British Raj came to an end. Below, we identify relevant literature that forms the basis of our investigation; highlight the methodology used which comprised a set of 30 in-depth interviews across the two research sites; present findings including four case-studies; and offer a discussion designed to extend the literature and our understanding of this recently emerging and significant theme.

LITERATURE REVIEW Comparative Entrepreneurship: A Framework The subject of IE has provided a major stimulus to research enquiry since Oviatt and McDougall’s (1994) conceptualization of international new ventures. The importance of younger entrepreneurial firms in the global economy is widely recognized (Young et al., 2003) as is their contribution to national and regional economic development in high technology and knowledge-intensive sectors (OECD, 1998). Despite the expanding literature in IE, numerous gaps remain, as identified, for example, by Zahra and George (2002), Young et al. (2003), Coviello and Jones (2004), Dimitratos and Jones (2005) and Zahra (2005). Of particular importance to this chapter are deficiencies in our understanding of the genesis of IE, and the effects of the institutional environment and economic geography on IE. This study on comparative IE in a global industry is particularly influenced by two recent papers, the first by Baker et al. (2005) on comparative IE in which the authors seek to develop a framework for comparing entrepreneurship processes across nations; and the second, by Fromhold-Eisebith (2004) on innovative milieu and social capital, especially relevant to the industry and country context of this chapter.

168

SHAMEEN PRASHANTHAM ET AL.

Building on the Shane and Venkataraman (2000) framework, the research by Baker et al. (2005) examines how and why entrepreneurial processes of opportunity, discovery, evaluation and exploitation differ across nations. The authors focus upon the antecedents of entrepreneurship, paying especial attention to the influencing roles of institutions and national cultures. Of particular relevance here is the opportunity exploitation stage of the entrepreneurial process where Baker et al. (2005) use ecological theory and economic geography to distinguish three representative types of ecology, namely, Regional Agglomerations, Broadly Developed Niches and Less Developed Niches. The literature on Regional Agglomerations is very extensive (from Marshall, 1920 onwards) with Silicon Valley and the ‘Third Italy’ being widely cited illustrations. Innovative regional agglomerations encourage new firm creation through easier access to specialized resources, such as venture capital (VC) and skilled labour, and state-of-the-art infrastructure and services (Tavares & Teixeira, 2006). Recognizing the stimulus to entrepreneurialism provided by regional agglomerations, Baker et al. (2005) also draw attention to potential challenges in such agglomerations such as an institutional ‘blind spot’ derived from their highly specialized nature; and the high and rapid returns and quick sales typically required by VCs. The contrast with the regional agglomeration is the Less Developed Niche, where entrepreneurs in developing and emerging economies lack access to specialized resources and institutional support. Family business groups (FBGs) are highlighted as one indigenous entrepreneurial response to these ‘institutional voids’ (e.g. Khanna & Rivkin, 2001), by providing a ‘gap-filling function’ to support new enterprise. According to Baker et al. (2005, p. 499), ‘FBGs are networks of many (usually small-scale) businesses that are linked together through kinship ties’ (see also Redding, 1990). While this strong kinship-based governance may stimulate entrepreneurship, subsequent growth may be inhibited because of the narrowness and myopia of kinship networks, and may be linked to national cultural differences. Interestingly for this paper, Baker et al. (2005, p. 501) mention Bangalore as an example of a ‘robust agglomeration’, and contrast this with most of India (to which one could add Pakistan) which resembles an ‘archetypal LDN’ (less developed niche). Overall, Baker et al.’s (2005) work is useful in understanding the social context of comparative entrepreneurship, and the categorization of ecologies provides a helpful way of distinguishing between India and Pakistan. But in general the article is weaker in its contribution to the ‘international’ dimension of comparative IE, a topic on which the present chapter has much to add.

Comparative International Entrepreneurship

169

Innovative Milieu and Social Capital In deepening our conceptual understanding, useful insights can be obtained from the work of Fromhold-Eisebith (2004), who addresses the notions of innovative milieu and social capital from a regional development perspective and focuses upon competitiveness within sub-national regions. These concepts have similarities with Baker et al.’s regional agglomerations and kinship ties in less developed niches. The notion of innovative milieu refers to ‘the complex network of mainly informal social relationships on a limited geographical area, determining y a sense of belonging, which enhance the local innovative capability through synergistic and collective learning processes’ (Camagni, 1991, p. 3). The notion of social capital refers to ‘the sum of the actual and potential resources embedded within, available through, and derived from the network of relationships possessed by an individual or social unit’ (Nahapiet & Ghoshal, 1998, p. 243). Fromhold-Eisebith (2004) articulates the differences between innovative milieu and social capital as follows: innovative milieu entail one-time or project-related interaction among heterogeneous actors with a focus on achieving change, in particular the commercialization of innovation. Social capital,1 on the other hand, involves everyday routines carried out by homogenous actors with a view to mastering capabilities and ensuring survival or stability. Her main point is that both of these aspects are relevant to the economic development of sub-national regions. Each set of actors provides benefits that have utility, particularly for smaller firms. The innovative milieu is a source of ‘unconnected resources and competencies’ and thereby ‘creative outcomes’ (Fromhold-Eisebith, 2004, p. 754). Social capital emanating from what is referred to as ‘firm communities’ provides support and advice to compensate for inexperience or deficiencies in everyday activities such as sales or hiring. An important point to note is that it takes access to a heterogeneous set of actors for certain benefits such as innovation outcomes to accrue. Thus for smaller firms in a regional agglomeration, a likely advantage over their counterparts in local niches is the novelty of information and opportunities that they have access to (McEvily & Zaheer, 1999). Of course, the likelihood that these will flow across a network is enhanced when social capital is built among actors (Inkpen & Tsang, 2005). Social capital fosters trust, thereby reducing barriers to exchanging and combining resources (Nahapiet & Ghoshal, 1998).

170

SHAMEEN PRASHANTHAM ET AL.

In the context of the development of international new ventures, the types of heterogeneous actors that are likely to facilitate entrepreneurial behaviour include companies, funding bodies, educational institutions, specialist suppliers, demanding customers and supportive policy-makers (Porter, 1998). This is particularly true of a global industry such as software. Investigating innovative milieu and social capital as complementary or redundant concepts of collaboration-based regional development, Fromhold-Eisebith (2004) places particular attention on the role of and relationships among actor groups. Because of this, discussion of key features of innovative milieu, particularly the institutional and infrastructural characteristics which are critical for creativity and innovation, is largely omitted. Similarly the notion of social capital as discussed by Fromhold-Eisebith (2004) does not discuss the important comparative dimension which derives in part at least from national cultural characteristics. Finally, considering these two core research papers, neither develops the genuinely ‘international’ dimensions of comparative entrepreneurial exploitation. These derive from the cross-border flows of ideas, knowledge, capital and human resources, which are vital ingredients in IE. These are important gaps to be explored in the current research.

METHODOLOGY Given the exploratory nature of the research, a set of in-depth interviews were undertaken in Bangalore and Lahore (Ghauri & Grønhaug, 2002). In each case the objective was to conduct interviews with a number of international entrepreneurs and to supplement these with interviews of other industry experts including academics, trade body officials and MNC managers. In addition, a large amount of secondary data from company websites and industry reports were studied. The use of multiple respondents and secondary data sources was a means to achieve triangulation (Miles & Huberman, 1994). The Bangalore fieldwork was undertaken in July 2006 by one of the authors. 16 interviews were conducted covering six software ventures, four MNC managers and four other experts including officials of the trade body for software, Nasscom. The ventures were identified through suggestions made by Nasscom. The objective was to interview highly innovative companies and the interviewees had all been recognized for meritorious innovation.

Comparative International Entrepreneurship

171

An interview guide was used to ensure some uniformity in the content of the interviews to facilitate data analysis. Interviews lasted on average for 90 min. In addition to the extensive field notes taken, the interviews were recorded. The Lahore fieldwork was undertaken in March 2007. A similar approach was adopted as in Bangalore, although in this instance all three co-authors participated in data collection. 14 interviews were undertaken, involving seven ventures and seven other industry experts. Again interviews lasted 90 min on average, and were tape-recorded and extensive field notes taken. Interviewees were identified using a snowballing technique; initial contacts were made by one of the authors with strong local networks. This was an inevitable approach given that this was an under-researched context. Data were analyzed in keeping with the approach advocated by Yin (1994). Thematic analysis of the content was undertaken first for each casefirm and then across cases. The additional expert interviews were then analyzed with a view to identifying the reiteration of or contradiction to themes identified from the interviews with the case-firms’ entrepreneurs. The general findings from the data are discussed initially, with characteristics of the local context (milieu) and the use of ethnic ties emerging as particularly salient issues. From the range of case studies, two from each setting that illustrate these differences particularly well were identified and are presented below.

FINDINGS: INDIA (BANGALORE) VS. PAKISTAN (LAHORE) The software sector in India is clearly very different to that in Pakistan in terms, for example, of longevity, overall size, worldwide reputation and the number of businesses of scale. But the countries themselves have common roots and British colonial backgrounds; and these differences and similarities provide a useful context for exploring the constituents of comparative IE. The Bangalore Software Industry Local Milieu According to Nasscom2 the Indian IT industry generated revenues of $30 billion during 2006–2007 – approximately 5% of GDP – of which nearly $24 billion (80%) was accounted for by international business. Nasscom

172

SHAMEEN PRASHANTHAM ET AL.

estimates that 1.3 million people are employed in this industry, claiming this to be ‘the largest pool of suitable offshore talent’, representing 28% of global offshoring labour. Yet, despite producing 501,000 engineering graduates in 2006–2007, India appears to face a manpower shortfall. To help bridge this gap, Nasscom proposes that a number of finishing schools for IT graduates be set up at leading institutions, as part of a wider industry–academia partnership agenda. Another challenge being addressed is that of setting up an ‘ecosystem for innovation’ to upgrade the level of knowledge-intensity of activities undertaken. Perhaps nowhere is the focus on capability upgrading perceptible as much as in Bangalore, which accounts for a third of India’s software export revenues. In relation to milieu characteristics, it seems fair to say that Bangalore has demonstrated signs of maturity – even though there is scope for improvement – on a number of key issues such as an entrepreneurial environment, FDI attraction, intellectual property rights (IPR) protection, access to capital and competencies, institutional support and reputation effects (Balasubramanyam & Balasubramanyam, 2000). Some of these issues are linked. For instance, Bangalore’s reputation effects have led to FDI attraction (and thereby the presence of leading MNCs) and access to VC, both domestic and international. Another key issue is the long-standing emphasis on education in public policy, which had led to a large pool of qualified engineers. It would be naive, however, to assume that there is no scope for improvement. It became apparent from the interviews that although Bangalore possesses a strong ‘soft infrastructure’ (Khanna & Palepu, 1997) in the form of, for instance, support from Nasscom, there were shortcomings in the physical infrastructure. Traffic congestion and prohibitive hotel rates were cited by respondents as impediments to the future progress of the Bangalore cluster. Ethnic Ties Significantly, compared to prior research findings from Bangalore (e.g. Prashantham, 2006) there was little mention of ethnic ties as being a critical resource. Rather, it appeared that the strong presence of Indian technologists in the US, and particularly in Silicon Valley (Saxenian, 2002) had resulted in the regular interflow of people, resources and opportunities between the two milieux. Besides anecdotal evidence of returnees from the US, the inter-milieu links were embodied in the work of a networking organization, The Indus Entrepreneurs (TiE). TiE had been founded in Silicon Valley as a

Comparative International Entrepreneurship

173

networking organization for technologists of South Asian origin. Chapters soon developed in several parts of the world, notably the US and India. The Bangalore chapter of TiE hosts regular events featuring guest speakers from Silicon Valley, many of them of Indian origin. For example, a recent such speaker was the noted VC, Vinod Khosla. Overall, however, it emerged from the interviews in Bangalore that Indian software ventures were not as heavily reliant on ethnic ties as they might have been once; institutionalized links with Silicon Valley appeared to exist and the Bangalore milieu had attained sufficient reputation effects (Prashantham, 2004) for ventures to be confident enough to engage with international markets without necessarily leveraging ethnic ties.

The Lahore Software Industry Local Milieu Lahore could not at present be described as a local milieu for the software industry. Statistics are hard to come by, but Pakistan Software Houses Association (PASHA) data indicate a total of 350 software houses based in Pakistan with revenues somewhere between $100 m and $200 m. The industry came into existence largely because of Y2K, and Lahore developed as the hub because the best universities were located there. The firms clustered in Lahore highlight a fragmented industry structure with no large companies (the biggest is NASDAQ-registered NetSol with about 600 employees); a few with approximately 50 employees; and the remainder micro-firms originating from universities in the US and Pakistan. The institutional structure is still in its infancy, although the Pakistan Software Export Board (based in Islamabad not Lahore) is gaining respect, as is PASHA and the Computer Society of Pakistan. Their influence and that of the companies themselves has been important in government education policy, resulting in a very large increase in investment from the early 2000s. Even so interviewees reported that the quality of graduates was lacking, in part because of the problems of getting IT faculty. The cost of labour is the major attraction, with direct labour costs 10–15% of those in the US; compared with India, one respondent cited direct labour costs in India of $40–45 per hour compared with $12–25 per hour in Pakistan. Telecoms infrastructure has improved significantly after earlier problems with bandwidth, although large additional investment is still required in data and cable Internet. Plans are underway for the establishment of a series of IT parks and incubators.

174

SHAMEEN PRASHANTHAM ET AL.

The regulatory environment was perceived as being weak, in terms, for instance, of protection of IPR, meaning that companies had to license IPR offshore; and the same was true for quality assurance. In terms of the operating environment for business, the interviewees contrasted the transparent environment in the US with the opaqueness of that in Pakistan. Related issues mentioned included ethics in business and payment problems. Overriding the internal challenges in developing a sustainable software industry was the ‘geopolitical’ situation and global market access problems. In particular, perceptions of US customers created anxieties towards outsourcing to Pakistan. This was particularly problematic when the internal market in Pakistan was still emerging, with a slow pace of computerization within government and its agencies. Ethnic Ties While much criticized, the Hofstede (2001) research on national cultures is a useful starting point for India–Pakistan comparisons. Scores for the individualism dimension rates India as 21st out of 53 countries, whereas Pakistan was 47th out of 53 (Hofstede, 2001, p. 215). The implication is that Pakistan is a highly collectivist society characterized by business behaviour traits such as: ‘keeping ethnic or other in-groups together supports productivity’; ‘the employee has to be seen in a family and social context’; ‘in business, personal relationships prevail over the task and the company’; ‘relationships with colleagues are cooperative for in-group members, hostile for out-group’; and ‘relatives of the employer and employee are preferred in hiring’ (Hofstede, 2001, pp. 244–245). Such notions were expressed by all respondents. Quotations from returnees from the US included the following: ‘Business is not like it is in the US – everyone relies on networks’; ‘personal networks are easy to build’ and ‘Pakistan society relies on favours’. Nearly all of the entrepreneurs interviewed studied at elite universities in the US (MIT and Stanford); but in raising finance, launching their businesses and finding initial customers (at least), they turned to ethnic Pakistanis. As mentioned below, Techlogix described the entrepreneurship consequences of this as leading to an (unsustainable) ‘friends/friends of the family’ business model. In a sense the Pakistani entrepreneurs were ‘within the US software milieu but not of it’. A meeting was held with the Lahore chapter of TiE, established, like the Karachi chapter, in 2000. The impression gained was that TiE was being widened out beyond its software roots, perhaps to stimulate membership.

Comparative International Entrepreneurship

175

A number of the entrepreneurs interviewed were TiE Lahore members; and Rozee (see below) in particular was very praiseworthy of the support that had been received in respect of credibility and access to business owners in Pakistan. However, other responses suggested an arm’s length approach to TiE, or reduced involvement because of limited benefits from the TiE association. One respondent suggested that the Organization of Pakistani Entrepreneurs of North America (OPEN), which was formed in the US in 2000, was perhaps an attempt to distance Pakistan from TiE.

CASE STUDY FINDINGS Skelta, Bangalore Skelta was founded in Bangalore in 2002 as a software product company. Skelta’s product orientation is relatively uncommon among Indian software companies, the vast majority of whom – including some of the best known ones such as Infosys and Tata – operate predominantly as services companies. Although some are sceptical about India’s prospects for producing genuinely world class software products companies, Skelta’s co-founders have been adamant in their self-belief that Skelta could achieve global success, based out of India. The founders include CEO Sanjay Shah, whose previous experience includes co-founding another software venture in India, iCode, which sought to offer enterprise-wide resource planning software for SMEs. Prior to this he co-founded Accel, a retailer of PCs, in Washington DC. Shah is a graduate of the prestigious Indian Institute of Technology (Mumbai) and holds a postgraduate engineering degree from the US. Another co-founder is Paritosh Shah, who is responsible for marketing and also has previous entrepreneurial experience in India and abroad, chiefly in the Middle East. Skelta’s flagship offering is known as Skelta BPM.NET. As evident from the name, this software application is focused on business process management (BPM), which involves the application of information technology to efficiently streamline and manage a variety of organizational processes. The second half of the product’s appellation indicates that the offering is built on Microsoft’s .Net platform technology, the underlying component on which software applications can be written for a Windows operating environment. The decision to ally its product technologically with Microsoft was a significant decision taken early on by the company founders. Microsoft

176

SHAMEEN PRASHANTHAM ET AL.

technology was seen as attractive given its widespread adoption by a range of companies across several countries. Thus, a company like Skelta that offers technology solutions that are built on a Microsoft platform stand a better chance of integrating with client companies’ extant Microsoft-based applications. From Microsoft’s viewpoint, such applications are of benefit too. Every time a Skelta product licence is sold, so is a Microsoft operating system licence. Not surprisingly, then, a core part of Skelta’s strategy has been to cultivate a deep relationship with Microsoft. Evidence of how seriously this relationship is taken is seen from the fact that the company has a separate function, alongside other conventional functions such as sales and finance, headed by a senior manager, which is referred to as a the Microsoft relationship function. It is also a measure of Skelta’s success in forging a cross-border, multi-faceted relationship with Microsoft. It is interesting to note, however, that Skelta was able to build a global relationship with Microsoft through efforts that began in Bangalore – its own backyard, as it were. A Microsoft manager in Bangalore who was responsible for forging relationships with technology partners such as Skelta notes that it was Skelta that had proactively approached Microsoft. Skelta’s technology impressed Microsoft as did the proactive approach of the top managers in seeking to gain visibility for Skelta. The Skelta–Microsoft relationship initially began at the local level through Skelta’s participation in Microsoft events and other promotional activities. For example, Skelta was invited to participate in Microsoft road shows whereby the product could be demonstrated to prospective clients in different parts of the country. Skelta’s CEO was invited to speak at a seminar organized for partner organizations of Microsoft in India on how Indian software products companies could achieve local and global success. Also, when Microsoft hosted an industry-wide workshop on the topic of innovation, Skelta was showcased as an example of an innovative company. Before long, the Indian subsidiary of Microsoft started promoting Skelta within the wider Microsoft network. To illustrate, in 2004 Skelta were invited to participate in the Microsoft Worldwide Partner Conference in Canada. The following year, Skelta was nominated by Microsoft India to be considered for an international innovation award at the Microsoft Conference in Boston, which it went on to win. In 2006, Skelta was among a set of internationally selected companies featured on Microsoft’s Vista promotion website, resulting in global visibility. By this stage, Skelta had acquired about 200 clients across a range of advanced economy markets including the US, UK and Canada.

Comparative International Entrepreneurship

177

Thus, the Skelta experience illustrates that, given relational capabilities of proactively establishing a valuable relationship with a significant player such as Microsoft, Bangalore provides local firms with access to international networks via the local environment. Clearly, the local milieu in Bangalore was able to attract the presence of large players like Microsoft, and this has greatly facilitated Skelta’s progress. As Skelta’s relationship with Microsoft has prospered so has the company, as evident from various distinctions achieved by Skelta in 2006. It was selected for a special award for innovation from the national trade body for software companies, Nasscom. It was recognized as one of Asia’s fastest growing technology companies by the trade publication Red Herring, which had famously recognized Google’s potential before many other industry observers had. Perhaps most significant of all, Skelta which had thus far been privately held, received its first round of VC amounting to $1.5 million from an Indian VC fund. Despite its considerable success, in the course of 2007, Skelta retained a Seattle-based management consultancy to facilitate collaboration with strategic partners and hired an ex-Infosys executive to head the company’s sales function for the US, based in Atlanta.

Liqwid Krystal, Bangalore Liqwid Krystal is a Bangalore-based e-learning company that was founded in 1999 by Anand Adkoli and a partner, Ramana Gogula (who has since left the world of technology for a successful career as a music director in the Hyderabad film industry). Adkoli, the co-founder who was interviewed, had gone to the US in the 1980s to pursue postgraduate studies in computer science. He stayed on in the US for over a decade where he worked for Oracle as a software architect in the server division. In parallel, he also developed a keen interest in computer education and authored several textbooks on programming with special reference to Oracle technology. Adkoli’s career with Oracle took him, in the 1990s, to a development centre in Melbourne and subsequently to a newly established facility in Hyderabad. By 1999, Adkoli felt the need for a change of direction in his career. He was strongly attracted to the prospect of returning to his native city of Bangalore and devoting more time to his writing interests. However, at this time a VC fund, Global Technology Ventures was set up in Bangalore in conjunction with the Bank of America. The VCs at this firm were scouting around for new business ideas. Attracted by the prospect of gaining

178

SHAMEEN PRASHANTHAM ET AL.

access to VC, Adkoli and his business partner proposed to set up an e-learning business, Liqwid Krystal. They were successful in raising $2 million from Global technology Ventures. In retrospect, it is without doubt that the Internet boom of that period compounded the VCs’ positive sentiments towards the business idea. Liqwid Krystal’s business proposition was based on the observation that whilst much information on computer programming was contained in the conventional medium of printed textbooks, much of the practical work was carried out in a completely different medium viz. through a keyboard and monitor of a PC. Consequently, it was often not easy for students to easily relate to what textbooks said about various aspects of programming. The founders of Liqwid Krystal believed that students could make a far better transition between theory and practice if they were given access to e-books whereby hyperlinks could be used to access the relevant software on which to write a piece of software code. This was a relatively simple but effective idea. Between 1999 and 2001 the company developed CodeSaw, a software product that integrated texts on programming with the facility for practical application by students. As for gaining access to textbook content, Adkoli used his publishing contacts in the US, given his own background as a textbook writer, to forge contractual relationships with such reputed publishers as Addison-Wesley and Thomson Learning. The basis for Liqwid Krystal’s revenues would be royalty payments whenever an e-book was sold from these publishers. However, the finalization of Liqwid Kystal’s offering coincided with a decline in the fortunes of the US information technology industry, triggering a decline in the derivative demand for computer education. The next couple of years were a period of struggle for the venture as it was locked into the publisher contracts. In 2004, it decided to terminate these contracts with the publishers. This allowed the company the flexibility to then negotiate contracts with a wider range of publishers that were less exclusive in nature. This meant that in addition to licensing content, Liqwid Krystal could provide such value-additions as assessment modules. Following this re-orientation of the company, Adkoli became conscious that he had ignored two rapidly growing markets for computer education, viz. China and India. He turned his attention to India first of all. His objective was to sign up universities as clients; the company would then provide an e-learning solution that every student could access at less than $10 per head per year. By 2006 the first contract was signed with an Indian university comprising 60,000 students. Subsequently, the company has forged a strategic partnership with the prestigious Indian Institute of Science

Comparative International Entrepreneurship

179

(IISc) in Bangalore. Computer students at an advanced level can study on-line via Liqwid Krystal’s learning platform called gyanX to obtain certification from IISc. The experiences of Liqwid Krystal suggest that their Bangalore base has provided them with access to VC, strategic tie-ups with prestigious academic institutions and a large market for on-line computer education. However, a complaint that Adkoli has – which is somewhat contradictory to Skelta’s positive experience – is that large companies, be they indigenous or foreignowned, are not as supportive of young ventures as they might be. As the company consolidates on its recent successes, the future plans involve making a foray into China with a view to tapping into the large education market in that county.

Techlogix, Lahore Techlogix is one of the largest and most-respected software companies in Pakistan. Registered in Bermuda, it employs 250 people worldwide, 30 of whom are in the United States, 170 in Pakistan where its Development Centre is located, and 40 in another Development Centre in China. The company operates a differentiated services model supplying major MNEs such as GE, Motorala and Shell, with 86% of revenues being generated in the US. The two co-founders are based in Pakistan but spend about half their time in the US. The co-founder of Techlogix who was interviewed – Salman Akhtar – has family in the US and took his BBS and MS degrees there before enrolling for a PhD at MIT. While at MIT he linked up with another Pakistani student with whom he had been at school, and together they established a company called The Technology Group in 1992/1993 to provide telephone banking software in Pakistan, growing the company from 2 to 14 employees and revenues of $100K by year four. New funding was required to grow the company in the US: Akhtar was unaware of the US VC business at this time and efforts were made to raise money in Pakistan; eventually most funding came from former Pakistani college room mates who had returned to Pakistan. $100K was raised and Techlogix was formed. Early in 1996 Akhtar linked up with a Pakistani whose company, Visage, located near Boston, operated in the driver’s licence market and required a ‘facts on demand’ software system. Techlogix established a Boston office to supply Visage, and for the next 2–3 years grew through business with Visage and by opportunistic customer extension;

180

SHAMEEN PRASHANTHAM ET AL.

the latter included business in the UK, which derived from a Pakistani connection. Looking for opportunities associated with the dot.com boom, Akhtar and his partner moved to California in 1999 and launched a new software product based on the latter’s PhD work at MIT from which the software was licensed. The product had a variety of applications including the car industry; and, while sales potential was substantial, long sales cycles were involved which presented cash flow problems for a small firm like Techlogix. Consequently the company returned to the services business and began its transition into enterprise software in California, with a specialization in commercial finance. An early breakthrough came with a successful tender for work with GE, based on an offshoring model; and GE (‘the godfather of outsourcing to India’) has remained as a leading customer. Gradually Techlogix developed a second-generation model based around five or six practice areas where they were world-class; and they began bidding for global vendor status as vendors consolidated on a worldwide basis. On this basis, the company was successful with Motorola, at a time when all their IT work was directed to Oracle or Infosys, and then with global players such as BTM and Shell. The basis of global competitive advantage is their differentiated services model, where differentiation combined with ‘best in the world’ expertise enables Techlogix to succeed against much larger competitors. In the early years of Techlogix, virtually all personnel were based in Pakistan; while Akhtar himself lived with his family in the US between 1997 and 2003 before returning to Pakistan. It was accepted that the Pakistan connection posed perception problems for some US customers; and the Development Centre in China was a customer-driven decision. This wholly owned Development Centre in Beijing is run by a Pakistani. As Techlogix globalizes in terms of customers and geography, access to global talent has become an issue: a recent high-level appointment was an American formerly employed by Infosys; and the company has hired five employees with Rhodes Scholarships. Efforts are being made to diversify the revenue base from a position where the US represents 86% of sales, and to include small as well as large customers. The attitude to business in Pakistan has until recently been ‘take it or leave it’, but Techlogix is now putting a serious sales team into the country. Similarly the composition of revenues has changed significantly – in the early 2000s, 80% of revenues were derived from application development, a figure which as at 2007 was down to less than 10%.

Comparative International Entrepreneurship

181

For the future Akhtar wanted to grow the company to a size of 1,000 people, which was considered to be feasible with a differentiated services business model. Interestingly, it was observed that at its current size, Techlogix would have attracted acquisition interest had it been an Indian company. As a Pakistani enterprise, the geopolitical situation was a barrier to the acquisition of a small firm like Techlogix. Commenting on the software industry in Pakistan generally, Akhtar described it as a being largely based on a ‘friends/friends of the family’ business model in which most companies were very small and could survive by ‘scratching around’. He argued that: ‘Outsourcing is not the basis for long-term competitive advantage. Outsourcing is a mechanism not a model’. By comparison, the Techlogix pitch was on specialization and complex solutions, using a ‘global delivery model’. Rozee.com, Lahore The founder – Monis Rahman – operates through a holding company, Naseeb Networks Inc., based in the US. Described as a ‘Silicon Valley new media company’, it employs 30 people in offices in Karachi and Lahore with Islamabad (Pakistan) to follow shortly (as of March 2007), and 2 people in America. This includes employees working in a Matchmaking site which had been launched in Pakistan in October 2003, but principally in Rozee.com which was established in August 2004 and is Pakistan’s #1 job website. Rahman studied at Stanford in the US and subsequently worked in California with Intel and ABB. His initial venture into personal entrepreneurship was in the US in 1997 with the establishment of a consulting business to site webcams in Day Care Centres. After one year’s experience the technology was acquired, and Rahman began to look for other opportunities. While he was exposed to a wide network, including VCs, through TiE, ideas for a new business engineered around the first received only limited funding support. While supporting his offshoring notion as a mechanism to reduce costs quickly, the VCs were asking for an office in India. Rahman returned to Pakistan in May 2003, but instead of providing offshore services for the US market, he identified a void for a matchmaking site targeting the Muslim community (his model was JDate a Jewish singles network). After six months working from home, the site was launched in October 2003. With ‘staggering numbers’ visiting the free site, Rahman spent $60K converting to a paid service which started on 19 April 2004. The investment was recovered in 2 months and this business continues to grow profitably as a global network. Currently 65% of revenues are generated in the US.

182

SHAMEEN PRASHANTHAM ET AL.

Rahman wanted to reinvest in other spaces, and, seeing first hand the problems of hiring in Pakistan, he launched Rozee as a job site in August 2004. Initially a ‘quick and dirty’ advertisement posting site, it was converted to a partially paid website in January 2007, paralleling the early usage of the Internet by major employers in Pakistan. There are plans to export the model from Pakistan to markets in the Middle East, and then globally, drawing on the experience from the matchmaking site. To date, growth has been completely organic, although Rahman has plans for various new growth avenues including a potential strategic alliance with Pakistan’s leading media house. Rahman described Pakistan as a ‘society based on favours’, where survival was not possible without networks. As a returnee, Rahman was recruited into the local chapter of TiE by the founder of the organization in Pakistan (who was also the founder of the leading Business School in the country – Lahore University of Management Sciences [LUMS]). Accepting that the TiE network in Pakistan provided Rahman with vital access to business owners and credibility, nevertheless, the nature of the TiE culture was perceived to be very different to that in the US, with suspicion replacing the transparency of US TiE. Market access internationally was considered as the biggest barrier to Pakistan software firms, and this applied particularly to links with multinationals where a real credibility problem existed. On the positive side, people were moving back to Pakistan, and the Vice-President Products for Rozee was formerly employed by AOL. As further evidence of optimism, Rahman cited the raising of VC funds in the US, and the recent acquisition of a neighbouring company – Cambridge Docs. – by a Chinese enterprise. Aside from market access, a major challenge was considered to be the limited number of IT graduates (20,000 per annum) being produced by Pakistan universities, and the limited proportion of these (2,000–3,000) who were of acceptable quality.

DISCUSSION: COMPARATIVE INTERNATIONAL ENTREPRENEURSHIP IN THE INDIAN SUB-CONTINENT SOFTWARE INDUSTRY The Indian and Pakistani software industries proved to be a useful context for exploring the topic of comparative IE; and similarly the notions derived from the literature were appropriate for comparative purposes. In general,

Comparative International Entrepreneurship

183

it would appear that Bangalore reflects characteristics of a regional agglomeration (Baker et al., 2005) which seems to relate to the innovation milieu notion (Fromhold-Eisebith, 2004), whereas Lahore depicts a less developed niche (Baker et al., 2005) that is heavily reliant on social capital (Fromhold-Eisebith, 2004) – and in particular on ethnic ties. In respect of India, the Bangalore milieu is recognized worldwide as providing a stimulating environment for creativity and entrepreneurship, illustrated in the company cases, for example, by the presence of world-class multinationals and local VC funds.3 The comparison with Pakistan could hardly be greater. The relatively small cluster of enterprises in Lahore depends greatly on ethnic ties and social capital, in part a replacement for weaknesses in the business and institutional environment. The advantages of Pakistani ethnic ties, consistent with the literature on bonding social capital (Putnam, 2000) include ease of information flow; access to seed capital; ease of building personal networks; overcoming national prejudices; and cross-border movement of businesses. However, there are disadvantages, associated with network closure, such as information redundancy; restricted opportunities; failure to ‘breakout’ of initial market; traditionalist aversion to high-tech entrepreneurship; and nepotism. Where our study goes beyond the literature (e.g. Baker et al., 2005) is by highlighting the role of cross-border linkages between milieux (see Fig. 1). Considering the theme of comparative IE, what is particularly interesting is to compare India/Pakistan–US and India–Pakistan relations as influences on entrepreneurship in the software industry. In different ways, bonds with the US are critical to IE in both India and Pakistan. Close inter-milieu links provide the opportunity to access complementary assets and business networks in the US. While Indians have had a huge influence on the development of Silicon Valley, their ties with Bangalore seem primarily to be based on hard-nosed business relations. But in relation to Pakistan, while the US milieu is critical for all aspects of the entrepreneurial process, the issue of closed networks may be a major barrier to long-term growth. As suggested earlier, ‘Pakistani entrepreneurs were within the US software milieu but not of it’. Thus, a major difference that emerged from our study between India and Pakistan is that, unlike in the case of Lahore, Bangalore’s linkages to Silicon Valley are increasingly transcending (often ad hoc) individual interpersonal ties to take on an institutionalized-like appearance facilitating a strong interchange of ideas, knowledge, capital and human resources through crossborder transactions and reverse migration (i.e. returnees from Silicon Valley).

184

SHAMEEN PRASHANTHAM ET AL.

Silicon Valley / Route 128 Indian

Bangalore

Ideas, knowledge, capital, human resources

Ethnic ties

Silicon Valley / Route 128 Lahore Pakistani

Ethnic ties Key: Outer boxes refers to an innovative local milieu (Silicon Valley / Route 128) or emerging local milieu (Bangalore). Refers to a cluster of firms but not a local milieu. Inner boxes. refers to strong ethnic ties. Refers to weak ethnic ties.

Fig. 1.

Local Milieu and Ethnic Ties: A Comparison of Cross-Border Entrepreneurial Linkages.

A vital implication of our observation pertains to the coordination and integration of the entrepreneurial processes of opportunity discovery, evaluation and exploitation across frontiers. Of theoretical interest is that this is an issue raised by Oviatt and McDougall (1994) that has received little subsequent attention. They described the ‘global start-up’ as a specific type of international new venture which ‘derives significant competitive advantage from extensive coordination among multiple organizational activities y [by] proactively acting on opportunities y ’ (Oviatt & McDougall, 1994, p. 59). Our research suggests that such coordination and integrating may actually commence at the pre-start-up phase, which resonates with their more recent definition of IE in terms of opportunities relating to ‘future goods and services’ (Oviatt & McDougall, 2005, p. 540; emphasis added). Useful managerial and policy implications follow. Learning from the Indian experience, policy-makers should undertake milieu-enhancing activities that yield potentially valuable networks to internationalizing new

Comparative International Entrepreneurship

185

ventures. Learning from the Pakistan experience, efforts should be made to leverage ethnic ties. This is a strategy that countries across the world, not just in Asia, are adopting, witness the GlobalScot initiative to allow Scottish business to forge links with expatriates. A key message for entrepreneurs is the importance of actively leveraging network relationships on a personal basis as well as via institutionalized arrangements such as TiE. As a concluding thought, we note also that geopolitical factors and sensitive India–Pakistan relations mean that the potentially beneficial complementarities that could stimulate software entrepreneurship in and between both countries are not being realized.

NOTES 1. It is interesting that she chooses to apply the ‘social capital’ label to only one of these two collective entities. Both notions have commonalities, particularly in relation to the emphasis on the importance of socially embedded interorganizational relationships within a local milieu (Inkpen & Tsang, 2005). Management researchers who apply social capital research in the study of entrepreneurial ventures (e.g. Davidsson & Honig, 2003) would perhaps argue that Fromhold-Eisebith (2004) is actually referring to two types of social capital viz. bridging and bonding social capital. Bridging social capital is associated with the benefits, such as innovation, arising from heterogeneous actors in an innovative milieu (McEvily & Zaheer, 1999); indeed, while analyzing the Silicon Valley cluster, Cohen and Fields (1999) make explicit reference to the social capital generated through interactions among such dissimilar actors as universities, policy-makers, companies and law firms. The stability and moral support arising from homogenous actors are associated with bonding social capital (Putnam, 2000). This of course may merely be a semantic matter. 2. These data can be found on the 2007 fact sheet for the Indian software industry published on the Nasscom website at http://www.nasscom.in/Nasscom/templates/ NormalPage.aspx?id ¼ 2374. 3. Yet it is far from being a fully developed innovative milieu comparable to Silicon Valley or Route 128. Gaps in the milieu relate, for example, to the quality of the physical infrastructure, the highly specialized nature of business activities and networks (Baker et al., 2005) and incomplete openness to trade and investment.

ACKNOWLEDGEMENTS We gratefully acknowledge financial support from the Department of Management, University of Glasgow (Lahore fieldwork) and the Carnegie Trust for the Universities of Scotland (Bangalore fieldwork).

186

SHAMEEN PRASHANTHAM ET AL.

REFERENCES Baker, T., Gedajlovic, E., & Lubatkin, M. (2005). A framework for comparing entrepreneurship processes across nations. Journal of International Business Studies, 36, 492–504. Balasubramanyam, V. N., & Balasubramanyam, A. (2000). The software cluster in Bangalore. In: J. H. Dunning (Ed.), Regions, globalization and the knowledge-based economy. London: Routledge. Buckley, P. J., & Ghauri, P. N. (2004). Globalisation, economic geography and the strategy of multinational enterprises. Journal of International Business Studies, 35, 81–98. Camagni, R. (1991). Innovation networks: Spatial perspectives. London: Belhaven Press. Cohen, S. S., & Fields, G. (1999). Social capital and social gains in Silicon Valley. California Management Review, 41, 108–130. Coviello, N. E., & Jones, M. V. (2004). Methodological issues in international entrepreneurship research. Journal of Business Venturing, 19, 485–508. Davidsson, P., & Honig, B. (2003). The role of social and human capital among nascent entrepreneurs. Journal of Business Venturing, 18, 301–331. Dimitratos, P., & Jones, M. V. (2005). Future directions for international entrepreneurial research. International Business Review, 14, 119–128. Fromhold-Eisebith, M. (2004). Innovative milieu and social capital – complementary or redundant concepts of collaboration-based regional development. European Planning Studies, 12, 747–765. Ghauri, P., & Grønhaug, K. (2002). Research methods in business studies. London: Prentice Hall. Hofstede, G. (2001). Cultures consequences. London: Sage. Inkpen, A. C., & Tsang, E. W. K. (2005). Social capital, networks, and knowledge transfer. Academy of Management Review, 30, 146–165. Khanna, T., & Palepu, K. (1997). Why focused strategies may be wrong for emerging markets. Harvard Business Review, 75(4), 41–49. Khanna, T., & Rivkin, J. W. (2001). Estimating the performance effects of business groups in emerging markets. Strategic Management Journal, 22, 45–74. Marshall, A. (1920). Principles of economics. London: Macmillan. McEvily, B., & Zaheer, A. (1999). Bridging ties: A source of firm heterogeneity in competitive capabilities. Strategic Management Journal, 20, 1133–1156. Miles, M. B., & Huberman, A. M. (1994). Qualitative data analysis. London: Sage. Nahapiet, J., & Ghoshal, S. (1998). Social capital, intellectual capital, and the organizational advantage. Academy of Management Review, 23, 242–266. OECD. (1998). 21st century technologies. Paris: OECD Publications. Oviatt, B. M., & McDougall, P. P. (1994). Toward a theory of new international ventures. Journal of International Business Studies, 25, 45–64. Oviatt, B. M., & McDougall, P. P. (2005). Defining international entrepreneurship and modeling the speed of internationalisation. Entrepreneurship Theory and Practice, 29, 537–554. Porter, M. E. (1998). Clusters and the new economics of competition. Harvard Business Review, 76, 77–90. Prashantham, S. (2004). Local network relationships and the internationalization of small knowledge-intensive firms. Copenhagen Journal of Asian Studies, 19, 5–26.

Comparative International Entrepreneurship

187

Prashantham, S. (2006). Foreign network relationships and the internationalisation of small knowledge-intensive firms. International Journal of Entrepreneurship and Innovation Management, 6, 542–553. Putnam, R. (2000). Bowling alone: The collapse and revival of American community. New York: Simon and Schuster. Redding, G. (1990). The spirit of Chinese capitalism. Berlin: de Gruyter. Saxenian, A. L. (2002). Transnational communities and the evolution of global production networks: The cases of Taiwan, China and India. Industry and Innovation, 9, 183–202. Shane, S., & Venkataraman, S. (2000). The promise of entrepreneurship as a field of research. Academy of Management Review, 25, 217–226. Tavares, A. T., & Teixeira, A. (2006). Multinationals, clusters and innovation. Does public policy matter? Basingstoke: Palgrave Macmillan. Yin, R. K. (1994). Case study research: Design and methods. London: Sage. Young, S., Dimitratos, P., & Dana, L.-P. (2003). International entrepreneurship theories: What scope for international business theories? Journal of International Entrepreneurship, 1, 31–42. Zahra, S. A. (2005). A theory of international new ventures: A decade of research. Journal of International Business Studies, 36, 20–28. Zahra, S. A., & George, G. (2002). International entrepreneurship: The current status of the field and future research agenda. In: M. A. Hitt, D. Ireland, D. Sexton & M. Camp (Eds), Strategic entrepreneurship: Creating an integrated mindset. Oxford: Blackwell.

This page intentionally left blank

REINTERPRETING A ‘PRIME EXAMPLE’ OF A BORN GLOBAL: COCHLEAR’S INTERNATIONAL LAUNCH Lisa Hewerdine and Catherine Welch The phrase ‘born global’ was coined by Michael Rennie of McKinsey in a study on high value-adding manufacturing exporters. In his article on ‘born globals’ in McKinsey Quarterly, Rennie (1993, p. 45) nominated Cochlear, an Australian company that produces hearing implants for the deaf, as ‘a prime example of a company that was ‘‘born global’’ ’. Rennie’s study has been very influential in academic circles – as has been his example of Cochlear, which has been repeated by Liesch, Steen, Middleton, & Weerawardena (2007), Madsen and Servais (1997) and Wickramasekera and Bamberry (2003). That Cochlear can be considered to be ‘born global’ (or an ‘international new venture’, to use the term introduced by Oviatt & McDougall, 1994) has not been questioned. Certainly, at first glance, Cochlear easily fits the ‘born global’ definition of ‘small, technology-oriented companies that operate in international markets from the earliest days of their establishment’ (Knight & Cavusgil, 1996, p. 11). The company was registered in 1983; in 1984 it established a US office to oversee clinical trials underway there, as well as launching trials in Europe; and in 1985 the US FDA approved its implant for sale, opening the door to the largest market for medical devices in the world.

New Perspectives in International Business Research Progress in International Business Research, Volume 3, 189–206 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1745-8862/doi:10.1016/S1745-8862(08)03009-4

189

190

LISA HEWERDINE AND CATHERINE WELCH

Thus, the company commenced international operations almost immediately upon establishment. Cochlear would seem to exemplify the born global in other ways too: it is the producer of a high-tech niche product that has successfully fended off competition from larger, better resourced multinational rivals (notably the US multinational 3M). Companies such as this have led to the question being posed: on what basis do such small niche players gain and sustain competitiveness (see, e.g., Almor & Hashai, 2004)? Are there in fact advantages associated with their rapid expansion, small size and thoroughly international outlook, unencumbered as they are by a legacy of domestic operations? In this chapter, we re-examine the case of Cochlear, arguing that to denote it a ‘born global’ is, at the very least, a simplification of its origins and early development. We offer a different interpretation of the years leading up to its establishment and international launch – the period of the company’s history neglected by previous accounts. We proceed by, first, presenting our analysis, and then explaining why the methodology on which it is based has led us to challenge existing accounts. We conclude by suggesting that the case raises some fundamental questions regarding the definition of born globals, and question whether ‘born global’ theories are suited to explaining the evolution of high-tech firms.

COCHLEAR’S EMERGENCE Development of a Prototype Cochlear’s first product, the 22-channel Nucleus implant, was the result of a research programme that has been dated back to 1967, when Graeme Clark, an ear, nose and throat (ENT) surgeon, commenced doctoral work on the electrical stimulation of the hearing nerve. Following the completion of his PhD in 1969, Clark was appointed the inaugural Chair in Otolaryngology at the University of Melbourne. When he joined the university in 1970, his primary objective was the practical application of his PhD research: namely, the development of a ‘bionic ear’, an electronic device that would stimulate the hearing nerve in the profoundly deaf. He realised early on that lack of resources would be one of his major impediments: across the road [from my office] the experimental research laboratory was in a disused hospital mortuary. When I looked at the mortuary my heart sank. It was dilapidated and bare. There was a stone table in the centre, but little else. The walls needed painting, and

Reinterpreting a ‘Prime Example’ of a Born Global

191

the light diffused poorly through the high windows. Anyway, I had no money to buy equipment even if the laboratory itself were satisfactory. (Clark, 2000, p. 54)

Professor Clark was able to obtain modest grants from his institution and a government medical funding body, but his needs proved greater, so he turned to donations from charitable trusts and public appeals through national telethons. He even resorted to collecting change from passers-by in Melbourne’s business district (Clark, 2000). He later estimated that between 1970 and 1977 expenditure on the bionic ear project amounted to A$576,770 (University of Melbourne, 1978b). In reflecting on these activities, he later commented, ‘I’m afraid that many scientists have to become entrepreneurs . . . if you have the drive to do the research you also need to drive to see that it’s funded . . . ’ (personal communication, 25 March 2008). With the money he raised, Clark formed a multidisciplinary team in partnership with Dr David Dewhurst, a colleague in Electrical Engineering. He rejected single-electrode stimulation of the hearing nerve in favour of a multiple-electrode approach that at the time was very controversial. Another decision made early on that gave the team a decisive technological advantage was the priority it gave to solving the problem of how to process speech (Clark, personal communication, 17 April 2008). By October 1977, Clark was able to submit the first of numerous patents for improvements to hearing prostheses. Although he successfully performed the first surgical implantation of his hearing prosthesis in a human patient in 1978, a first step towards demonstrating its commercial potential, the implant was still crude: ‘The components were barely acceptable surgically and reduction in the size of components was necessary. Development of the [s]peech processor was needed’ (University of Melbourne, 1978a). This development phase would require even more money than the expenditure to date, as well as advanced bioengineering and commercial expertise. The search for a commercial partner therefore commenced in earnest in 1977–1978. Two serious contenders for this role emerged. 3M Australia initiated contact with Professor Clark in October 1977 after having read about his ‘bionic ear’ in a local newspaper. At the same time, the team was in touch with staff at Telectronics, an Australian manufacturer of pacemakers. Telectronics was a subsidiary of Nucleus, which in this period was Australia’s most internationally successful medical technology group. Both 3M and Telectronics negotiated agreements with the University of Melbourne in 1977 in order to evaluate the technology. However, in September 1978, 3M decided that it would back a rival American singleelectrode device that was also under development.

192

LISA HEWERDINE AND CATHERINE WELCH

Meanwhile, the reaction from Telectronics and from the owner of Nucleus, Paul Trainor, was from Clark’s perspective disappointingly ‘slow’ (Clark, 2000). Geoffrey Wickham, then a director of Telectronics and well known to Dewhurst, recommended to the Board that the project be taken on, but without success. Wickham conveyed to Clark his opinion that it would be ‘difficult, if not impossible, to gain funding for a project in which the ultimate cost could not be reasonably predicted’ (personal communication, 9 May 2008). Yet contacts between the two organisations continued. Two key R&D staff1 at Telectronics, Michael Hirshorn and David Money, both took trips to Melbourne to evaluate the device in 1978. While Money was impressed by the technological advance that the cochlear prosthesis represented, he was also convinced that if implanted in a human it would fail (which indeed it subsequently did). His assessment was that in order for the device to be made more reliable, a complete redesign would be necessary (interview, 7 May 2008).

Government Involvement Given the lack of interest from industry, it was suggested to Clark that he should appeal to the Australian Government, which at that time had boosted funding for industrial research and development with the objective ‘to bridge the gap between research and the successful translation of resultant inventions into commercial production’ (Cooley, 1978). Accordingly, he wrote to the Secretary of the Department of Productivity in July 1977, explaining that ‘[t]he University at the moment has been approaching firms in Australia to see if they are interested in developing this device and I believe it would help if the Australian Government could provide financial assistance’ (Clark, 1977). The Department of Productivity dispatched a senior official to Melbourne to investigate, and he sent back a highly positive report on the potential of the innovation. The Department also sought advice from Trainor, who served on government boards throughout his career and was well known to policymakers. Clark was asked to submit a request for funding accompanied by a costed development plan. He did so in March 1978, but waited months for a response as the Government was waiting to see if any commercial partners would step forward. Crucially, the ‘bionic ear’ was seen to fit the Government’s policy objectives of nurturing a high-technology base in the country and countering the trend of Australian innovations being ‘lost’ to foreign companies. The policy was predicated upon the Government

Reinterpreting a ‘Prime Example’ of a Born Global

193

providing the bridge between research and industry: ‘To reverse the trend of declining international competitiveness of Australian industry, the Commonwealth [Government] recognises that it must take a leading role by undertaking, with the assistance of industry, major projects . . . ’ (Department of Productivity, 1979). In December 1978, two key meetings were held at the University of Melbourne. At one, the Department indicated that it was prepared to award Clark a A$400,000 ‘public interest’ grant that would fund the next phase of development, which was agreed would consist of the implantation of two more patients and the completion of an international market survey and cost estimates within a nine-month period. The Government would provide the ‘risk capital’ for development on the condition that the device would be commercialised in Australia. At the second meeting, Paul Trainor declined the role of commercial partner. He explained that he would not be prepared to take the project to his Board unless he could defend its chance of success as being much greater than 50% – and at the present time, he estimated that the potential for success was less than that threshold (University of Melbourne, 1978a). Trainor later recalled that he was the only one at this point to understand how costly and lengthy the ‘bionic ear’ project would be – a judgement backed up by his head of R&D, David Money. It was upon Trainor’s recommendation that the Government planned for a market and cost analysis to be undertaken before a final decision was made to proceed with the development of the device. As Trainor later recalled, ‘To me . . . , having reviewed over 300 such projects before . . . , mainly in medical fields, the initial response is always: ‘‘let’s explore it more carefully’’ ’ (Gibbs, 1999). Despite Trainor’s pessimism, Telectronics was nevertheless able to assume a role in the project. The Government announced a A$400,000 public interest grant on 31 January 1979, and on the following day advertised for expressions of interest from companies to undertake a market survey and development cost plan. Telectronics was one of seven companies invited to bid and one of only two which proceeded to submit a tender. Telectronics, in which Trainor had acquired a controlling stake in 1967, was at that stage among the top five in the world market for cardiac pacemakers. It had manufacturing facilities in the US and France, as well as a worldwide network of sales offices and distributors. The Telectronics tender pointed out that the company’s track record provided it with considerable experience in the development and marketing of implantable electronic devices: ‘we feel we are admirably suited to carry out this project as it is of a generally similar nature to the work which we have to do for any introduction of new products of our own’ (Trainor, 1979).

194

LISA HEWERDINE AND CATHERINE WELCH

The Department of Productivity concurred with these claims in its evaluation of the Telectronics bid. The company had in-house R&D and manufacturing capabilities for implantable devices, which meant that the development cost plan would be conducted ‘by people who are likely to be aware of the manufacturing problems likely to be encountered’ (Schultz, 1979). The company’s marketing expertise was assessed as being similarly well qualified. Telectronics had recently completed a similar marketing plan for its latest invention, the bone growth stimulator. As Trainor later summed it up, Telectronics was the ‘logical’ choice (Gibbs, 1999) – and arguably the only company at that time in Australia with the required industrial and international experience – so it came as no surprise when it won the government contract, enabling it to commence work in July 1979.

Public Interest Grant: Stages 1–3 Even before the outcome of the tendering process was announced, David Money and a colleague, Chris Daly, had been pondering the challenge of the re-engineering of the device. They devised a radically simplified design that represented a considerable inventive step. The original prototype had almost 50 integrated circuit chips; two coils, one of which was for the provision of power and the other of data; and a speech processor that had to be connected to a minicomputer for the essential programming for the individual patient. The Telectronics design contained a single integrated circuit, one coil for both power and data, and envisaged a diagnostic and programming unit that could be controlled by a home computer. The new design, on which the development cost plan and subsequent development work was based, was to deliver not just improved reliability but greater capability as well: notably, it was able to stimulate 21 sites in the inner ear rather than the 10 sites stimulated by the original prototype. In this first phase of the public interest grant, a steering committee was set up, chaired by an assistant secretary from the Department of Productivity. As contractors, the University of Melbourne and Telectronics sat on the committee and were obliged to report to it on their activities. Their final reports were submitted in December 1979, on the basis of which the Department decided that a second phase of funding could be recommended. In total, four rounds of government funding took place from 1979 to 1985, totalling about A$4.7 million, and the steering committee continued to meet until the end of 1985. In Phase 2 Telectronics moved from being a consultant to providing the much-needed hardware improvements, with

Reinterpreting a ‘Prime Example’ of a Born Global

195

their first task being the miniaturisation of the device. The involvement of Telectronics in this development work was very much based on the advice provided to the Government by Clark and Dewhurst, who felt that the company’s background in implantable devices was vital in producing a hearing implant that could withstand the corrosive nature of the body (Clark, personal communication, 17 April 2008). At this stage, Nucleus itself became the contractual party rather than its subsidiary Telectronics. Telectronics management and its French shareholder, the pharmaceutical company Synthelabo, felt the cochlear implant project did not fit with the company’s plans and the potential return on investment was seen as being too far in the future (Gibbs, 1999), so they consented to the project’s move into Nucleus. Stage 3 of government funding, which was publicly announced in September 1981, differed from earlier rounds in that not only was it the largest (A$2.127 million), but it was the first time in which more funding was now going to Nucleus as the commercial partner rather than to the basic research being conducted by the University of Melbourne team. In a contract signed with a government agency, Nucleus committed to deliver a prosthesis that was ready for clinical trials. Conditional upon this agreement was the negotiation of an exclusive worldwide agreement between Nucleus as licensee, and the Government and University of Melbourne as joint licensors. The licensing agreement, signed in December 1982, provided the licensors with a joint share of royalties amounting to 5% of sales of the first 12,000 units and 2% thereafter. Nucleus won the right to sub-license, although any move to manufacture offshore would require prior approval from the licensors. At this stage Trainor set up an expert ‘tiger team’ within Nucleus (although initially it consisted of just three staff), to conduct the bioengineering for the implant (Epstein, 1989). As Clark (2007) recalled: Paul Trainor’s ‘tiger team’ included some staff from my team to the university. We both considered that this transfer of personnel and know-how would be the best and quickest way to achieve a commercial outcome. I was also committed to ensuring that our research would continue to focus on further support for the industrial development.

Clark paid tribute to the fact that Trainor reciprocated this collaborative spirit: ‘[he] strove hard to ensure that the commercial effort dove-tailed with our research, and as time went on, together we developed such a rapport in explaining our work at symposia that it seemed to me we performed like a ‘‘Punch and Judy’’ show’ (Clark, 2000, p. 147). In 1982 the ‘tiger team’, in close collaboration with Clark’s group, succeeded in delivering a device,

196

LISA HEWERDINE AND CATHERINE WELCH

based on Money and Daly’s design, that Clark successfully implanted in pre-clinical trials involving six patients. The Nucleus implant betrayed its origins: it derived some of its components, technology and design (such as the use of titanium encapsulation and a hermetic electrical feedthrough) from the Telectronics pacemaker.

Public Interest Grant: Stage 4 The Government never guaranteed funding beyond the existing phase. In Phase 2, an accounting firm was contracted to produce detailed economic costings on the project. Its report recommended that the Government continue to support the project given that ‘the risks associated with the project at its present stage of development are too high for prudent, though entrepreneurial private enterprise management to commit the total funds required for the project’s completion while faithfully discharging normal obligation to shareholders’. In 1982, at the end of Phase 3, by which time the implant was ready to undergo clinical trials, Nucleus made a submission to the Department (now reorganised and renamed the Department of Science and Technology) requesting that government funding be continued because ‘at this stage in Nucleus’ growth it is not possible to fund such a high-risk project’ (and in fact, the Nucleus Group had suffered a loss in 1982 due to problems with the development of a second-generation diagnostic ultrasound). The report stressed the long lead times before medical products become profitable, and the fact that the implant was yet to receive regulatory approval. An independent consultant commissioned by the Government agreed with Nucleus’ assessment of the degree of risk, as did the venture capitalists who were approached but declined to participate in the project. The Department of Science and Technology reached an agreement that it would fund 75% of Phase 4, with Nucleus contributing the remainder. While it declined to fund the clinical trials for the cochlear implant, Nucleus emphasised in its submission to the Government that the company would be eager to assume responsibility for commercialisation at the end of Phase 4: ‘This project is particularly important to Nucleus Limited because of its potential for diversification within the high-technology health care field’ (Nucleus, 1982). Nucleus was a suitable candidate ‘because of its international experience in the health care industry’. By this stage Nucleus could boast that it had successfully commercialised three Australian innovations: several generations of the cardiac pacemaker, a bone growth stimulator and a diagnostic ultrasound. The cochlear project very much

Reinterpreting a ‘Prime Example’ of a Born Global

197

fitted the company’s strategy, which was ‘to innovate and develop products which represent as far as possible an advance in applied technology rather than a mere duplication of products already existing in the market’ (Prospectus, 1980, p. 8). This strategy had been successful: when the holding company went public in 1980, it had four operating companies, 700 staff, revenues of A$41.8 million, and worldwide reach, with Australia only representing 23% of total sales (Prospectus, 1980). Nucleus had signalled early on that if commercialisation proceeded, it would at this point set up a separate subsidiary to produce and market the implant. This was consistent with the ‘Nucleus philosophy’ – implied by the company’s name – ‘of the centre being in control and attracting additional cells to it’ (Nucleus Prospectus 1980, p. 7) that had been followed with previous innovations (such as the ultrasound, which had been spun off into the subsidiary Ausonics). The reasoning behind a separate company structure was that ‘it gives the independence required for rapid, unencumbered growth which allows it to benefit from the expertise and resources of the parent, Nucleus’ (Nucleus, 1982). It would also enable the possibility of attracting additional shareholders – potentially, it was suggested, even the Government and the University. Nexus Biomedical (soon renamed to Cochlear Pty Ltd) was set up as a wholly owned subsidiary company of Nucleus in 1983. In 1985, after years spent pursuing investment partners, Trainor succeeded in attracting a A$3million capital injection by an Australian-based syndicate led by Western Pacific Investment Company. At the same time as establishing a separate corporate identity, the new company remained very much part of the Nucleus Group. The existing international presence of the Group proved most useful in Japan, where the local Telectronics office was able to contribute contacts and market knowledge (and in fact, within a few years Cochlear was to buy out Nihon Telectronics). Cochlear benefited from other supporting linkages within the Nucleus family; for example, Telectronics undertook all the sterilisation of the cochlear implants. In 1983 clinical trials commenced in Australia and the US – putting the Australian team behind 3M, which had already commenced trials of its single-electrode system. The two surgeons who performed the first operations in the United States were attached to medical centres that had been in contact with Clark’s team for some years, and they made trips to Melbourne in preparation for the surgery. The cochlear team had been making regular trips to the US and Europe in order to sign up clinical centres. Clark provided introductions to surgeons he knew, Nucleus staff cold-called other potential contacts, and an audiologist in the United States

198

LISA HEWERDINE AND CATHERINE WELCH

(who had worked for a period at the University of Melbourne) was engaged as a consultant to evaluate potential centres and make contact with key decision-makers in the field. The task of signing up US centres for the trials was made more challenging by the fact that, as the 1979 market survey by Telectronics concluded, the Melbourne University group ‘is not well known or understood elsewhere’ and ‘[a]n Australian device entering an American market in competition with American devices must start at a disadvantage’ (Telectronics, 1979). However, Hirshorn (who by this stage had moved to Nucleus) was able to report to the steering committee three years later that medical centres in the US were both aware and favourably disposed towards the Melbourne University implant – a testament to the sustained effort that Nucleus had invested over this period to raise the profile of the Australian invention. Preparing for Approval and Sales Nucleus had also spent years preparing for a successful application to the US Food and Drug Administration (FDA). FDA approval was recognised as critical in the 1979 Market Survey, since ‘the United States is likely to be the first major market for cochlear implants’ – yet in Telectronics’ experience with pacemakers, the hurdle for foreign companies was greater than that for US applicants. The Telectronics market survey team first met with FDA officials in September 1979, and even at this early stage the team impressed upon Clark the need for his documentation and trials to meet FDA requirements. Regulatory approval was a team effort requiring multiple trips to Washington over a period of several years, as well as the use of local advisors. Just as the technological development of the implant owed much to the existing resources of the Group, so it was when it came to preparing the groundwork for regulatory approval. The Group had an existing association with a corporate lawyer in the US, who had extensive experience in regulatory matters and who had advised on earlier FDA applications. The pre-market approval application that was submitted in 1984 ran to 3,000 pages and involved meticulous preparation; as Hirshorn later explained, some documents were even ‘re-written . . . to correct the subtle differences between Australian and American English’ (Hirshorn, n.d.). It was in this period in the lead-up to FDA approval that Cochlear’s competitive strategy was devised. Trainor was very much involved in the discussion on pricing. Based on his experience over many years, Trainor argued strongly that that it would be essential to start with a high price.

Reinterpreting a ‘Prime Example’ of a Born Global

199

While he was not supported by his own marketing team, after much discussion his view prevailed – ensuring that the new company made its first profit to the group by 1986. David Money, Cochlear’s first CEO, and his colleagues carefully implemented a marketing approach that would exploit what they perceived as 3M’s weaknesses: Experienced clinicians were recruited rather than salesmen on the assumption that potential customers wanted information and training . . . Specific training courses were instituted for audiologists and otologists (ear surgeons). Emphasis was placed on a team approach to patient care and maximum clinical benefit . . . A spirit of selling a philosophy rather than a product was developed. (Hirshorn, n.d.)

In 1985, weeks before the FDA announced its approval of the Nucleus implant (a year after its approval of 3M’s device, which was first to market), Paul Trainor sent a letter to a newly appointed Government minister that provided a history of the cochlear project. He concluded that the project reflects what can be done here in Australia when Government, University and Industry really cooperate. It may well prove to be a case study that should be duplicated by others for the advancement of the well-being of society. (Trainor, 1985)

Trainor’s confidence in the success of the cooperation was vindicated by the fact that by 1987, Nucleus could claim that Cochlear ‘is now well established as the market leader worldwide in cochlear implants’ (Nucleus Annual Report, 1987, p. 7) – a position which it has maintained ever since. In 1989, 3M admitted defeat and exited the industry (Van de Ven, Polley, Garud, & Venkataraman, 1999) with Cochlear purchasing its rival’s implant technology in a move to reassure patients and maintain confidence in the fledgling industry.

METHODOLOGICAL NOTE: THE VALUE OF ARCHIVAL SOURCES Our historically grounded narrative of Cochlear’s origins is very different to the ‘born global’ portrayal of a start-up company that achieved instant internationalisation. The difference can in part be attributed to one of method. Our account of the Cochlear case in the previous section relies substantially on the papers of Professor Graeme Clark. As well as preserving copies of his correspondence with the Department of Productivity and Nucleus, his manuscript collection includes an almost complete set of the steering committee minutes from 1979 to 1985. While the collection is

200

LISA HEWERDINE AND CATHERINE WELCH

obviously weighted towards Clark’s own activities, it also contains the transcript of an interview he arranged with Paul Trainor in 1999, government minute and briefing papers, and internal documents from Cochlear on the early international marketing efforts of the implant. Thus, insights into the motives and actions of all three parties (government, industry and university) could be gained from the collection. The chapter was also reviewed by former senior staff at Telectronics and Nucleus: Robert Foot, Mike Hirshorn, David Money, Anne Simmons and Geoffrey Wickham. Our archival analysis stands in contrast to previous studies, which were either largely survey-based (McKinsey & Company, 1993) or dependent on interviews with a later generation of Cochlear executives who were not present in the early years prior to the company’s formation (Liesch et al., 2007). Unlike these methods, archival analysis is well suited to excavating events and processes that occurred over a longer period of time. In the case of Cochlear, it can be argued that previous accounts, in limiting the temporal boundaries of the case to 1983, have in fact skewed their conclusions. Yet careful drawing of a case’s temporal boundaries is an important part of the ‘casing’ process (for a discussion, see, e.g., Dubois & Gadde, 2002). In the following section we explain the implications of bringing in the history neglected by previous accounts of Cochlear’s evolution.

DISCUSSION AND CONCLUSION Cochlear Pty Ltd was established by Australia’s largest and most internationally successful designer, manufacturer and marketer worldwide of biomedical technology. In 1983, when it founded Cochlear, the Nucleus Group had 800 staff (300 of which were located offshore), 27 subsidiaries and global reach, with international sales reaching 80% of total sales. The new subsidiary was able to benefit from the parent company’s culture of ‘internationalism’ (as Trainor put it), technical expertise, leadership and financial backing. The spin-off company was, moreover, following a template for the commercialisation of new technologies that the Group had already refined with its pacemaker, bone growth stimulator and ultrasound. In particular, Cochlear could not have emerged had it not been for the international outlook and technological expertise provided by Telectronics. Nucleus had, moreover, experienced the risks as well as successes of commercialisation. Robert Foot (2004), who was Nucleus company secretary for many years, has observed that Nucleus – and hence Cochlear – was able to learn from the losses incurred in the development of

Reinterpreting a ‘Prime Example’ of a Born Global

201

the second-generation diagnostic ultrasound: ‘the experience so gained, ensured that, when Nucleus made the decision to take on the Cochlear project, it would be in a much stronger position from the outset to have the reserves in terms of people and capital to ensure success’. Viewed solely from the perspective of this corporate heritage, Cochlear could be described as a case of successful corporate entrepreneurship. As Trainor emphasised, the cochlear implant would not have been commercialised without the backing of Nucleus at a time when other prospective investors had cried off: ‘but for Nucleus’ persistence in going ahead, there would have been no viability, no ongoing research and no commitment’ (Gibbs, 1999). Moreover, Cochlear was by no means the Nucleus Group’s first corporate venture, and was able to benefit from the Group’s accumulated experience of commercialisation – including, in the case of Ausonics, collaboration with government and research institutions. This is not to underestimate the challenges that the new subsidiary faced. The parent company’s resources were stretched at that time and the existing operating companies were understandably not prepared to subsidise the new venture. The international networks that had been established over the years for devices such as pacemakers and ultrasound did not extend to ENT surgeons. Hirshorn (2003) recalled that when he was sent to set up Cochlear Corporation in the US in 1984, Trainor gave him a cheque for A$10,000 and the latitude to establish the new company ‘wherever you can be successful’: from that base, he had to build the US subsidiary. However, Cochlear was much more than a bold internal corporate venture. As Trainor also acknowledged, the company was the outcome of a three-way partnership bringing together university, government and industry. Ultimately, Cochlear had not one but three organisations involved in its founding: the Department of Productivity, University of Melbourne and Nucleus. Nucleus had declined to underwrite the project through the development stage so government funding, lasting for five years, was indispensable. But even before the Government stepped in to provide the ‘risk capital’, Professor Clark had managed to keep his project alive despite the constant danger of running out of funds. This required more than just brilliance and vision as a researcher: he was also tireless in his quest for financial support. Nor was his role over once the basic research had been completed and Nucleus took over as project manager. Clark provided unstinting research support throughout the development phase, gave expert advice to Nucleus when required, and provided contacts the company needed to set up clinical trails and recruit medical centres. Market acceptance of the device would not have been possible without the

202

LISA HEWERDINE AND CATHERINE WELCH

recognition Clark’s research established, despite initial scepticism, in the research community. The ‘bionic ear’ was always intended by the three parties involved – the founding partnership of government, university and industry – to be targeted at international markets. This point was stressed by Clark in his initial submission to the Government in 1978, in which he wrote: ‘A large export market is obviously available assuming manufacturing can be undertaken in Australia’ (University of Melbourne, 1978b). Export potential was one of the criteria for the Government’s public interest grant. The Government’s support, which remained steadfast during the many years of development, was based on its assessment that the project would provide export earnings, local employment and the growth of related industries. Trainor, too, instantly recognised that the cochlear prosthesis would have to compete internationally, so when the Department of Productivity first contacted him about the project his immediate reaction was to start making calls to his contacts offshore (Gibbs, 1999). By the time the Cochlear subsidiary was established in 1983, it was thoroughly global in its outlook and ambitions. Cochlear’s internationalisation was not, however, as rapid as may first appear when it is considered that the international marketing plan for the ‘bionic ear’ was prepared in 1979. As Trainor repeatedly emphasised to others involved in the project, the development and commercialisation of a high-technology product is unavoidably a long-term endeavour: ‘We knew from our previous experience in developing implantable devices that the whole project, from prototype to accepted medically improved (FDA) implant, would take at least ten years’ (Gibbs, 1999). In other words, Cochlear’s seemingly ‘instant’ internationalisation required years of effort – not just to overcome the technological hurdles that Clark’s group and later the ‘tiger team’ faced, but to undertake the time-consuming and delicate process of establishing relationships with medical experts and regulatory authorities in international markets prior to product launch. Thus, by the time Cochlear was established in 1983, it was a somewhat old ‘new venture’. Trainor’s estimate of 10 years (which does not even include Clark’s years of basic research) is a reminder of the length of time it can take to bring a high-technology innovation to market. This period of time does not fit easily into the timetable that researchers have prescribed for a born global – whether this be two, three or even eight years. This suggests that in order for a high-tech firm with a new innovation to qualify as a born global, development of the innovation needs at least in part to have preceded the

Reinterpreting a ‘Prime Example’ of a Born Global

203

formal establishment of the company. The pre-founding development process that the product has already undergone, including the first steps offshore, may well have a substantial effect on the firm’s subsequent internationalisation path. The influence of the innovation history predating the firm can also be detected in Oviatt, McDougall, Simon, and Shrader’s (1993) account of Heartware, a ‘born international’ medical equipment company. Heartware was incorporated in 1988 and began its first marketing efforts in 1989. The firm’s sole product, an electrophysiology system, had been developed by two researchers in Holland and was in-licensed from the University of Limburg. The state of the system at the time of the licensing agreement, the relationship between Heartware’s CEO and the researchers, and the scientific credibility and networks of the original inventors were aspects of the development phase of the innovation which affected Heartware’s path upon foundation. It meant that when Heartware was founded, the electrophysiology system already had a history which imprinted itself on the company. In Heartware’s case, this pre-founding history left it in a much more precarious position than Cochlear: it had no financial backer or parent company and no in-house technical expertise; relations between the Heartware CEO and the researchers were not close and well developed; and awareness of the Limburg system in the US medical community was not high. Thus, to exclude this product development phase from analysis is to overlook factors relevant to the firm’s internationalisation and performance. Our reanalysis of the Cochlear case suggests that to denote it a ‘born global’ or ‘international new venture’ is problematic, and a denial of the protracted process of evolution that took place. Was the Cochlear venture really established in 1983, or should its genesis be traced back to the mid1960s, when Nucleus was formed?2 Or 1979 when the Government and Nucleus first joined the university effort? Or in 1981, when Paul Trainor set up the tiger team in Nucleus? Or was Cochlear really only truly ‘born’ when it gained independence from its parent upon its IPO in 1995? Certainly, any account of Cochlear should not exclude or downplay the period, prior to the existence of the company, in which the research and development of its first invention took place, or the role of the corporate parent and its organisational collaborators. Yet once these events are taken into account, it becomes less meaningful to claim Cochlear is a ‘born global’. Cochlear is of course just one case – albeit an iconic and high profile one. Single cases can nevertheless provide tests of existing theories and form the basis for further theoretical development. Cochlear forms what George and Bennett (2004) have termed a ‘most-likely’ case: one that is expected to fit

204

LISA HEWERDINE AND CATHERINE WELCH

well with an existing theoretical explanation. Given that the posited explanation did not survive this test, we would suggest that refinement is required when it comes to the emerging theory of born globals. It is not simply that existing methods of classifying companies as born globals – of operationalising key constructs – may need to be recalibrated. More fundamentally, it would appear that there needs to be a closer examination of the nature of the venture being examined. If the venture is defined at the level of the innovation itself, the company entity – and the timing of its founding – becomes less central. Further consideration of the unit of analysis under investigation may therefore be required. In addition, our analysis has pinpointed a broader issue surrounding the high-tech nature of the venture. The Cochlear case suggests that high-tech firms – hitherto considered to be more likely to be born global – may have protracted and complex pre-establishment histories, given the lengthy and costly period of product innovation that they need to undergo. The seemingly rapid expansion of a high-tech venture may in fact be the result of much of its technological (and even commercial) development occurring prior to the legal founding of the company. Greater understanding of the linkages between the innovation process, company founding and internationalisation would therefore seem to be warranted. Thus, while our interpretation of the Cochlear case has not provided definitive answers, it has raised important considerations for theoretical development and future empirical research.

NOTES 1. Given that Paul Trainor disliked the word ‘manager’ and it was not used in the Nucleus Group, the term has also been avoided in this chapter. 2. A record of the exact date when Trainor established Nucleus has not survived, and depending on the source he puts the founding date as occurring between 1964 and 1966.

ACKNOWLEDGMENTS We would like to express our gratitude to Professor Graeme Clark for granting us access to his manuscript collection at the National Library of Australia, and for generously making himself available to answer our questions and comment on a draft chapter. Staff at the National Library of

Reinterpreting a ‘Prime Example’ of a Born Global

205

Australia, particularly Bronwyn Ryan, were also extremely helpful. Given that Nucleus no longer exists (it was taken over in 1988) and its founder Paul Trainor died in 2006, tracking down the history of the company has required the assistance and cooperation of others. We would especially like to acknowledge Geoffrey Wickham and Matthew Trainor for providing us with our starting point; and Robert Foot, Mike Hirshorn, David Money, Professor Anne Simmons and Geoffrey Wickham for commenting on chapter drafts and sharing their recollections. The interpretation, of course, remains our own, as do any remaining errors.

REFERENCES Almor, T., & Hashai, N. (2004). The competitive advantage and strategic configuration of knowledge-intensive, small- and medium-sized multinationals: A modified resourcebased view. Journal of International Management, 10, 479–500. Clark, G. (1977). Letter to Sir Alan Cooley, Secretary, Department of Productivity, 18 July, Box 10, Papers of Graeme Clark MS8696, National Library of Australia. Clark, G. (2000). Sounds from silence: Graeme Clark and the bionic ear story. St Leonards: Allen & Unwin. Clark, G. (2007) Boyer lecture 4: Imagination becomes a reality, ABC Radio National, 2 December. Available at http://www.abc.net.au/rn/boyerlectures/stories/2007/2084245.htm. Accessed on 23 January 2008. Cooley, A. (1978). Letter to R. Marginson, 21 February, Box 10, MS8696, National Library of Australia. Department of Productivity. (1979). Background and objectives of public interest projects, document prepared for industry briefing 10 April, Box 11, Papers of Graeme Clark MS8696, National Library of Australia. Dubois, A., & Gadde, L.-E. (2002). Systematic combining: An abductive approach to case research. Journal of Business Research, 55, 553–560. Epstein, J. (1989). The story of the bionic ear. South Yarra: Hyland House. Foot, R. (2004). Ausonics, 9 May, In: M. P. Trainor (Ed.), Paul Murray Trainor, unpublished. George, A. L., & Bennett, A. (2004). Case studies and theory development in the social sciences. Cambridge, MA: MIT Press. Gibbs, M. (1999). Interview with Paul Trainor, October 1999, Box 54, Papers of Graeme Clark MS8696, National Library of Australia. Hirshorn, M. (2003). Letter to Matt Trainor, 3 August, In: M. P. Trainor (Ed.), Paul Murray Trainor, unpublished. Hirshorn, M. (n.d.), Cochlear: High tech internationalization of an Australian invention against a large USA multinational, Box 19, Papers of Graeme Clark MS8696, National Library of Australia. Knight, G. A., & Cavusgil, S. T. (1996). The born global firm: A challenge to traditional internationalization theory. Advances in International Marketing, 8, 11–26.

206

LISA HEWERDINE AND CATHERINE WELCH

Liesch, P., Steen, M., Middleton, S., & Weerawardena, J. (2007). Born to be global: A closer look at the international venturing of Australian born global firms (August). Australian Business Foundation. Madsen, T. K., & Servais, P. (1997). The internationalization of born globals: An evolutionary process? International Business Review, 6(6), 561–583. McKinsey & Company. (1993). Emerging exporters: Australia’s high value-added manufacturing exporters. Melbourne: Australian Manufacturing Council. Nucleus. (1982). Bionic ear: Implantable hearing prosthesis, report to Department of Science and Technology, Nucleus, 21 December, Box 14, Papers of Graeme Clark MS 8696, National Library of Australia. Nucleus. (1980). Prospectus. Nucleus. (1987). Annual Report. Oviatt, B. M., & McDougall, P. P. (1994). Toward a theory of international new ventures. Journal of International Business Studies, 24, 45–64. Oviatt, B. M., McDougall, P. P., Simon, M., & Shrader, R. C. (1993). Heartware international corporation: A medical equipment company born international, Part A. Entrepreneurship Theory and Practice, 18(2), 111–128. Rennie, M. W. (1993). Born global. McKinsey Quarterly, 4, 45–52. Schultz, P. (1979). Letter to J.A. Clark, Australian Industrial Research and Development Board, 22 May, Box 16, Papers of Graeme Clark MS 8696, National Library of Australia. Telectronics. (1979). Market study: The commercial feasibility of cochlear implantation, Box 12, Papers of Graeme Clark MS 8696, National Library of Australia. Trainor, P. (1979). Tender for development cost plan and marketing survey for the cochlear implant, 3 May, Box 11, Papers of Graeme Clark MS8696, National Library of Australia. Trainor, P. (1985). Letter to Barry Jones, Minister for Science, 14 August, Box 17, Papers of Graeme Clark MS8696, National Library of Australia. University of Melbourne. (1978a). Minutes of meetings, University of Melbourne, 7 December 1978, Box 11, Papers of Graeme Clark MS 8696, National Library of Australia. University of Melbourne. (1978b). A submission from the University of Melbourne to the Department of Productivity. Box 10, Papers of Graeme Clark MS 8696, National Library of Australia. Van de Ven, A. H., Polley, D. E., Garud, R., & Venkataraman, S. (1999). The innovation journey. Oxford: Oxford University Press. Wickramasekera, R., & Bamberry, G. (2003). Exploration of born globals/international new ventures: Some evidence from the Australian wine industry. Australasian Journal of Regional Studies, 9(2), 207–220.

MYTHS IN MICROFINANCE Roy Mersland and R. Øystein Strøm ABSTRACT Microfinance – the provision of financial services to the poor – is high on the public agenda. We discuss and evaluate three myths regarding microfinance based on new data from rated microfinance institutions (MFIs). The first myth is that an efficient MFI needs to be shareholder owned; second that its governance should first and foremost address the potential conflict between owners and managers; and third that MFIs are drifting away from their poorer customers towards serving the wealthier. The data do not support any of these myths. We conclude that microfinance is a viable business model.

INTRODUCTION Microfinance is high on the public agenda following the UN Year of Microcredit in 2005 and the Nobel Peace Prize awarded to Mohammad Yunus and Grameen Bank in 2006. Christen, Rosenberg, and Jayadeva (2004) report that 500 million persons are served by microfinance, mostly through savings accounts. In the 2006 Halifax meeting, the Microcredit Summit celebrated reaching the milestone of 100 million borrowers. Nevertheless, microfinance still only reaches a fraction of the world’s poor (Robinson, 2001; Christen et al., 2004). Donors and policymakers are

New Perspectives in International Business Research Progress in International Business Research, Volume 3, 207–227 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1745-8862/doi:10.1016/S1745-8862(08)03010-0

207

208

ROY MERSLAND AND R. ØYSTEIN STRØM

therefore active in issuing recommendations to increase service outreach (Helms, 2006; C-GAP, 2004, 2006). In this chapter we argue that several of these recommendations are built on myths, not facts. The myths are that shareholder ownership structures are more efficient than the non-profit structures popular within microfinance; that microfinance institutions (MFIs) require Western-style governance structures; and that MFIs gradually tend to migrate towards serving customers in higher income brackets in order to pursue financial sustainability. New evidence casts doubt on the truth of these myths. In particular, large-scale global data from rated MFIs warrant a reconsideration of the accepted truths. We analyse these popular myths in both theoretical and empirical investigations. The chapter proceeds as follows. The first section explores the myth that an MFI needs to be a shareholder-owned firm (SHM) in order to achieve sustainability. The second section addresses the myth that an MFI needs to implement Western-type governance mechanisms in owner–manager relationships. We stress that a horizontal bank–customer relationship must complement the vertical owner–manager relationship, and regressions show that the horizontal relationship is the more important. The third section deals with mission drift, that is, that MFIs leave their poor customer segment in favour of wealthier customers in order to improve their financial results. The fourth section discusses the overall findings and concludes.

MYTH 1: THE NEED TO BE A SHAREHOLDER-OWNED FIRM Does the type of ownership of an MFI make a difference to its performance? Here we build upon Mersland and Strøm (2008a), who investigate this question on a sample of rated MFIs. MFIs constitute a diverse set of ownership types. The firm may be privately held by shareholders, a sub-unit of a larger bank, a state bank, a member-owned cooperative, or a nongovernmental organisation (NGO). Since the NGO Prodem in Bolivia was transformed into the SHF Banco Sol in 1992, it has been argued that an evolutionary organisational process that transforms non-government MFIs into SHFs is required (Pischke, 1996). Accounts of successful transformations (Fernando, 2004) and guidelines on how to transform have been published (Ledgerwood & White, 2006; White & Campion, 2002). The arguments are that SHFs can be regulated by banking authorities, accept deposits, provide a larger range of better quality services, be independent

Myths in Microfinance

209

from donors, attract private equity capital, and benefit from superior corporate governance because they are privately owned. The claim is clear: SHFs perform better than NGOs. Nevertheless, the issue of transformation has so far created more discussion than action. Of the thousands of NGOs, only about 43 have transformed into SHFs (Hishigsuren, 2006). However, an alternative hypothesis may be that SHFs and NGOs do not perform differently because they may use the same business model to compete and serve customers in the microfinance market. In fact, different ownership forms are common in the banking and insurance industries (Mayers & Smith, 1983; Hansmann, 1996). In mature bank markets where different ownership types co-exist, researchers find little evidence to suggest that ownership type influences operational efficiency (Altunbas, Evans, & Molyneux, 2001; Crespi, Garcia-Cestona, & Salas, 2004; ESBG, 2004). In a recent large European study, Iannotta, Nocera, and Sironi (2007) found that investor-owned banks have higher profitability, but have higher operating costs than non-investor-owned banks. In historic terms, pro-poor banking has generally been dominated by mutual and non-profit ownership, not by investor ownership (Cull, Davis, Lamoreaux, & Rosenthal, 2006; Hansmann, 1996). The question remains: why do policymakers advocate a shareholder charter for MFIs? Thus, our question is whether the assumed superiority of SHFs compared to NGOs holds in microfinance markets. Does the type of ownership matter? Mersland and Strøm (2008a) investigate the claimed double bottom line of the MFI, both its outreach to poor customers and its financial performance. Here, we focus upon financial performance, which gives an indication of the MFI’s sustainability. This should prove to be the sharpest test of the claim that the SHF is the superior organisational model for microfinance. We perform tests on rated MFIs. Being rated means that the MFI opens the books to international, specialised rating agencies (the reports are publicly available at www.ratingfund.org). Reports from the following five rating agencies are included: MicroRate, Microfinanza, Planet Rating, Crisil, and M-Cril. The methodologies applied by the rating agencies have been compared and no major differences in how they assess MFIs have been found. All five agencies are approved official rating agencies by the Rating Fund of the Consultative Group to Assist the Poor (C-GAP). The fact that the MFIs in the sample are rated means a certain selection bias in that the data are skewed towards the better-performing MFIs. However, this is an advantage in our comparative analysis since much background ‘‘noise’’, such as very small MFIs or development programmes

210

ROY MERSLAND AND R. ØYSTEIN STRØM

without the intention of applying microfinance in a business-like manner, have been filtered out. This allows for a more realistic comparison between ownership types. The data sample is 68 SHFs and 132 NGOs. The rating reports making up the database date from 2000 to 2006, with the vast majority from the last three years. The rating reports contain financial information for up to four years. As required, all numbers in the dataset have been annualised and converted to USD using prevailing official exchange rates. The rating agencies differ in terms of the information they make available in the reports, thus a different N on different variables and in different years is reported. We implement a test with the logit regression method in order to predict the organisational type for a binary dependent variable, showing 1 for the NGO and 0 for the SHF from a set of background variables. The explanatory variables are chosen based on their likelihood to differentiate between the SHF and the NGO, and are based on Schreiner’s (2002) framework to compare the performance of MFIs. We have four proxies to measure financial performance differences between SHFs and NGOs: debt/ assets ratio, operating expense ratio, portfolio at risk, and return on equity (ROE), which shows the equity cost. We also include return on assets (ROA), which in many ways summarises the aforementioned proxies. The outreach performance differences are the average loan amount, indicating the depth of outreach, and the number of credit clients, indicating breadth. If the myth of better efficiency in SHFs is to be upheld, we expect the financial performance proxies to be better for an SHF than an NGO. The rating reports contain information from up to four years. Reports on estimation for all the four years are presented in Table 1. The omnibus w2 ð8Þ test is a Wald test for the null hypothesis that all coefficients in the equation are zero. We can reject this hypothesis in all specifications. The Nagelkerke R2 measure shows how much is explained. This measure gives values that are usually much smaller than those in linear regression models. Therefore, the statistic shows satisfactory results. Also, the percentage of cases correctly classified indicates that the overall regression performs well. Hence, the power of our statistical model is strong. Since SHF is coded 0 and NGO 1, a positive sign indicates a higher probability of detecting the NGO, while a negative sign indicates a higher probability of detecting the SHF. Thus, from Schreiner’s framework, we expect NGOs to have lower average loans compared to SHFs, while financial performance measures should show negative signs indicating better performance in SHFs. Specifically, the ROA should be negative.

211

Myths in Microfinance

Table 1.

Logit Calculations of Organisational Predictions. Year 0

–1

–2

–3

Debt level Operating portfolio expense ratio PaR 30 ROE Average loan amount Credit clients ROA Total voluntary savings Constant

0.612 3.769 6.793 0.344 0.000 0.000 3.753 0.000 0.260

0.949 2.407 7.935 0.455 0.000 0.000 6.695 0.000 0.246

1.482 1.868 5.217 0.232 0.000 0.000 2.912 0.000 0.996

2.128 2.615 7.745 0.964 0.000 0.000 0.954 0.000 1.201

Observations Classified correctly (%) Omnibus w2 (8) test Nagelkerke R2

148 79.1 0.000 0.399

144 79.2 0.000 0.377

136 79.4 0.000 0.369

91 79.1 0.000 0.449

Note: Years 0 to –3 when the binary variable ownership type contains SHF coded as 0, and NGO coded as 1. Year 0 is the rating year, while year –1, –2 and –3 are the previous years.

Table 1 shows that our measure of depth of outreach, average loan amount, is not significant in any regressions. Overall, few significant results are obtained, indicating that it is difficult to pick out the type of ownership from the Schreiner (2002) dimensions. The negative debt level (year –3) and the positive operating portfolio expense ratio (years 0 to –2) have the correct signs according to this hypothesis, as do the results for PaR30. However, the ROA is everywhere positive and also significant in year –1. This is contrary to the hypothesis that SHFs are more financially efficient compared to NGOs. Thus, although costs and risk are higher in the NGOs, NGOs seem to develop a business model that has an ROA equivalent to or better than the SHFs. Several robustness tests are performed. For example, the debt/equity ratio replaces the debt level in one specification, while in others, ROE and ROA are alternately removed in order to avoid potential correlation between the two. We also check if the various explanatory variables are significantly different in the NGOs and the SHFs. None of these tests upset the results above. We simply cannot differentiate between NGOs and SHFs on the basis of financial results. This also extends to outreach, the MFI’s ability to serve poor customers. The SHFs and NGOs are equally good or bad.

212

ROY MERSLAND AND R. ØYSTEIN STRØM

It seems to us that Dichter (1996) is correct when he observes that many NGOs involved in microfinance are driven by the same economic rationality as any other bank. The results are not necessarily surprising. Ownership theories do not predict a clear preference for one type of ownership in the microfinance market (Mersland, forthcoming). This is further supported by findings in general banking markets, as well as the pro-poor banking history, indicating that mutual and non-profit ownership can compete successfully with investor ownership. The recommendation for NGOs to transform into the SHF organisational form is premature, and seems to be built on a myth.

MYTH 2: WESTERN VERTICAL GOVERNANCE Mersland and Strøm (2008b) carry the ownership type analysis further by asking which governance mechanisms are effective in MFIs. We look at the relationship between firm financial and outreach performance on the one hand, and on the other, characteristics of top management (board and CEO), ownership type, and the external governance mechanisms of bank regulation and competition. Besides Hartarska (2005), no rigorous empirical study has looked at the governance of MFIs. The data that Hartarska (2005) use are limited to Eastern Europe, while ours include rating reports on 278 MFIs in 61 countries throughout the world for a period of up to four years. The governance problem in a bank has more to do with the horizontal agency relationships with customers than the traditional Jensen and Meckling (1976) vertical relationship between owners and management. The problem is perhaps exacerbated in microfinance since the borrower has little or no collateral and credit history, and the depositor has generally low education and little knowledge of the bank. Moreover, the regulatory ability of local regimes is generally low. Issues such as closeness to the customer and mutual trust are paramount under such conditions. However, for the bank it is also important to make sure that lent capital is returned. Hence, typical microfinance loans are small and of short maturity, the typical customer is a woman, and group lending is often the loan methodology. The governance myth is that an MFI needs to have the same governance structure as those recommended for Western developed firms, for example in OECD (2004). This implies setting up governance systems to mitigate the

Myths in Microfinance

213

vertical agency problems by aligning top management with owners’ profit goal and putting controls into place (Rock, Otero, & Saltzman, 1998; Otero & Chu, 2002; Helms, 2006). We believe that these recommendations are well and good, but not sufficient to understand the governance problems of MFIs. For one thing, it detracts from the horizontal agency problems often encountered in banks (Adams & Mehran, 2003). A second concern is that most MFIs are not shareholder owned and therefore have no clear profit motive, but need a profit in order to survive. At the same time, many SHFs are dominated by social investors who may stress their mission to reach out to poor customers as much as NGOs. These aspects motivate a study of governance in MFIs, and to include horizontal issues in the analysis. Both the weight placed on horizontal agency issues and the broader data differentiate this study from Hartarska (2005). We study firm performance using ROA as the main financial performance variable and average loan as the main outreach performance measure. In the following we use arguments for financial performance, since this is the most studied in the literature. The explanatory variables are vertical, horizontal, and external governance mechanisms. The vertical variables measure the traditional owners and management agency relationship, and the horizontal variables measure the bank and customer relationship. In general, we expect the vertical governance mechanisms to be of less importance in an MFI compared to the horizontal. Level of competition and regulation are the external mechanisms studied. The vertical governance mechanisms include board characteristics. The variables are the duality of the CEO and chairman, the presence of international directors, an internal board auditor, and board size. Since CEO/chairman duality is likely to reduce the owners’ control, we expect this to be negatively related to ROA, although Brickley, Coles, and Jarrell (1997) cannot confirm this for US data. Oxelheim and Randøy (2003) show that international directors have a positive impact upon firm performance, and we expect likewise here. To have an internal auditor reporting directly to the board is a common recommendation in order to improve control in the industry (Steinwand, 2000), and is therefore included in our analysis. Board size is negatively associated with firm performance in previous studies (Yermack, 1996; Eisenberg, Sundgren, & Wells, 1998). To the vertical board variables we add ownership type. The findings we report in Myth 1 lead to the prediction that the SHFs and NGOs show equally strong financial performance. Moreover, Crespi et al. (2004) find

214

ROY MERSLAND AND R. ØYSTEIN STRØM

that Spanish savings banks have comparable or better financial results than shareholder-owned competitors. The horizontal governance mechanisms include the female CEO and the loan methodology. Considering the gender of the CEO and the loan methodology to be governance mechanisms is novel and unexplored in the literature. However, since agency costs in microfinance are as much to do with the customer–bank relationship as the owner–manager relationship, we consider it important to identify the effect of these horizontal governance mechanisms. Most of the microfinance loans are extended to women. The average female proportion out of 109 firms reporting this variable is 74% in our sample. This being so, we believe a female CEO is better at understanding the MFI market than a male. Therefore, we expect the female CEO to be positively related to firm performance. The loan methodology is a dummy variable showing 1 if the loans are mainly made to individuals. Microfinance is credited for discovering, or rediscovering, group lending. In group lending, the group as a whole is generally responsible for repayment. The ‘‘social capital’’ inherent in the internal control among group members acts as a form of collateral (Tirole, 2006, pp. 180–181), and the screening task for removing bad credit risks from the group resides with group members (Armendariz de Aghion & Morduch, 2005). From a theoretical standpoint, we therefore expect group lenders to show better firm performance. However, former studies in microfinance indicate that individual lenders have the highest financial return, while group lenders reach out to the poorest customers (Cull, Demigu¨c-Kunt, & Morduch, 2007). Therefore, we expect the sign of the loan methodology to be uncertain. Competition and bank regulation belong to the external governance mechanisms. In general, higher product market competition should lead to lower firm performance (Hart, 1983; Schmidt, 1997), although the higher competitive pressures could lead some firms to improve through cost reductions. Petersen and Rajan (1995) show that banks lose the rent they earn on long-term relationships when competition increases. Thus, we expect competition to be negatively associated with firm performance. Bank regulation gives MFIs the opportunity to accept deposits and gives easier access to ordinary capital markets, which should lead to better firm performance. However, many MFIs are not regulated, and the cost of regulation may outstrip the benefits. One cost item can be the deadweight costs of reporting and the extra security that regulation entails. Another is the economies of scope in combined lending and deposits, which only seem to be positive in very large banks (Berger & Humphrey, 1997). Moreover,

215

Myths in Microfinance

Hartarska and Nadolnyak (2007) find that regulation has no direct influence upon MFI performance. Thus, we expect the effect of regulation upon MFI performance to be uncertain. In order to test the various relationships, we use the same data described under Myth 1. However, this time the dataset is enlarged to include a total of 278 MFIs from 61 countries. We use panel data estimations for the task. Since some of the explanatory variables are categorical, the random effects method is the correct one to use (Woolridge, 2002). We report results for ROA, portfolio yield, and operating expenses in Table 2. The Wald test (Greene, 2003) shows satisfactory overall test results. We run regressions, first including vertical governance mechanisms and then leaving them out. In separate regressions, entering vertical governance mechanisms individually does not improve the results. The regressions also contain control variables that are often used in similar research (Cull et al., 2007). The HDI shows the country’s state of development, which is a country control variable.

Table 2. Firm Performance Explained by Vertical, Horizontal, and External Governance Mechanisms. Random Effects Panel Data 3SLS Estimation Spanning the Period 1998–2007. Return on Assets Constant CEO/chairman duality International directors Board size SHF Female CEO Individual loan Competition Regulation Urban market Portfolio at risk (30) MFI experience Firm size Human dev. Index

0.196 0.034 0.002 0.001 0.002 0.027 0.001 0.006 0.033 0.006 0.07610 0.001 0.0161 0.001

Wald F (sign.) Firm years

0.001 435

Portfolio Yield 1

5

0.0365 0.012 0.007 0.026 0.024 0.1291 0.002 0.0281 0.056

0.504 0.1255 0.019 0.002 0.016 0.002 0.0945 0.019 0.1245 0.0975 0.079 0.003 0.0255 0.4745

0.000 628

0.000 434

0.420

Operating Cost 1

0.025 0.1021 0.01610 0.07210 0.0841 0.019 0.002 0.012 0.3425

1.3201 0.1135 0.03410 0.000 0.005 0.067 0.052 0.011 0.028 0.043 0.12510 0.002 0.0831 0.41210

0.0745 0.0971 0.014 0.024 0.029 0.024 0.0055 0.0711 0.177

0.000 632

0.000 426

0.000 617

0.464

1.4131

Note: A raised number indicates a significant result, with the number representing the significance level, e.g., the superscript 5 means a significant result at the 5% level.

216

ROY MERSLAND AND R. ØYSTEIN STRØM

First, look at the vertical governance mechanisms. We see that board characteristics and ownership type have little impact upon firm results. The CEO/chairman duality is significant in portfolio yield and operating costs, and international directors only for the latter. In both regressions they are of the ‘‘wrong’’ sign, implying either higher portfolio yield or higher operational expenses. The international director result is a surprise. International directors should bring more competence and better network capabilities. On the other hand, international directors are presumably less knowledgeable about local conditions than local directors. Perhaps they bring in a culture of higher costs. A tie-in to local information networks is probably more important to microbanks than international control. Furthermore, we find that ownership type (SHF) is nowhere significant. This confirms the Mersland and Strøm (2008a) result that SHFs are neither more profitable nor more efficient than NGOs, which concurs with international evidence (Crespi et al., 2004). It turns out that the horizontal mechanisms of female CEO and individual loan explain firm performance better than the vertical mechanisms. This confirms the importance of the bank–customer relationship in microfinance. A female CEO improves ROA and reduces operational costs. More individual loans bring down portfolio yield and operational costs, but do not significantly affect the ROA. Thus, in order to improve its financial results the MFI should have a CEO who is close to its market, and this should increase the importance of individual loans. The individual loan result is counter to a widely held belief that group lending is necessary for success in microfinance. We cannot confirm this, therefore our results are in line with Cull et al. (2007). Presumably individual lending can be organised more efficiently, and the individual collateral available gives the bank a better risk assessment opportunity than group lending. Whether group lending results in lower outreach to poorer customers, as found in Cull et al. (2007), remains unanswered. Nevertheless, the results for the horizontal governance mechanisms confirm the general hypothesis that they are important in the MFI. The external governance mechanisms of competition and bank regulation are significant in the portfolio yield regressions. In addition, the competition variable has the expected sign in every regression, and has a significance level of about 15% in the ROA and operating costs regressions. Regulation and competition are important in holding portfolio yield down. Thus, the external governance mechanisms are important. Taken together, we find that governance matters for MFIs. In addition to the vertical relationship between owners and managers, we investigate the

Myths in Microfinance

217

role of horizontal relationships between the bank and its customers, as well as the external governance mechanisms of competition and bank regulation. Arguably, when customers have a shorter credit history and less experience with financial institutions than in Western developed economies, and when MFIs operate in less regulated environments, MFIs should value horizontal relationships more highly. We find that this is the case, and conclude that the Western-style governance myth must be moderated for MFIs.

MYTH 3: MFIS DRIFTING AWAY FROM THEIR MISSION The microfinance industry is coming of age, and with this comes claims that it is straying from its mission to serve the poor (Dichter & Harper, 2007). An MFI’s mission is to provide banking services to the poor, that is, to lend very small amounts of money to very poor borrowers. In this chapter we ask if MFIs are staying true to their mission, or if there is a ‘‘mission drift’’ away from the very poor client segment. The ‘‘microfinance schism’’ (Morduch, 2000) in the mission drift debate has several variants. Woller, Christopher, and Warner (1999) and Woller (2002) hold that MFIs suffer from mission drift when they start providing loans to wealthier community members. Another position is that MFIs are more concerned about making money for their owners than staying true to their mission. Rhyne (1998) and Christen and Drake’s (2002) view is that a more commercialised microfinance industry is better able to serve the poorest members of the community, since their profit motive leads them to be more efficient and more willing to seek out new markets for their loan products. Mersland and Strøm (2008c) take the view that mission drift is defined relative to the poor customer segment. Serving the poor can be done commercially or non-commercially. A recent survey of 704 MFIs by the Microbanking Bulletin (2007) reveals that 41% are not financially self-sustainable and rely on donor support to remain afloat. Thus, financial viability is a major concern. Nevertheless, the risk of mission drift has been high on the agenda ever since the Bolivian MFI Prodem was commercialised and transformed into Banco Sol in 1992 (Rhyne, 1998). In a qualitative study, Hishigsuren (2007) follows one MFI in Bangladesh using archival, survey, and interview data from different stakeholders. This important case study concludes that the MFI shows no statistically

218

ROY MERSLAND AND R. ØYSTEIN STRØM

significant mission drift measured on depth, quality, and scope of outreach to poor clients. The only relatively large cross-country study where mission drift is also a concern is by Cull et al. (2007). From a sample of 124 MFIs in 49 countries they find little evidence of mission drift and conclude that MFIs can stay true to their mission, even when aggressively pursuing financial goals. The high-tempered debate and lack of cross-country surveys covering a large number of MFIs indicates the need for this study. Our data are well suited to the mission drift issue. In this chapter we use observations of 290 rated MFIs from 61 countries. Third-party organisations carry out the ratings and outside organisations subsidise part of the costs involved. At each rating, four years of data are obtained at best. Thus, the maximum number of observations is 1,160. The ratings are carried out for the period 2001–2007, which means that we have data from 1998 to 2007, with most data from the period 2001 to 2005. Thus, we may study mission drift for the sample of MFIs both in calendar time, relative to the year rated, and relative to the MFI’s first year in microfinance. In general, the monetary variables are converted into USD amounts at the going exchange rate. From the purchasing power parity theorem of international finance (Solnik & McLeavey, 2004), the conversion into USD implies that local inflation has been adjusted for. In addition, we filter out the worst outliers, that is, the average loan is defined to lie between USD50 and USD10,000. First, what do we mean by ‘‘mission drift’’? In our view, mission drift refers to the migration of MFIs away from the very poor customer segment and into other market segments. The argument behind mission drift is that commercialisation, that is, the drive for better financial results, will drive the MFI away from its poor customers. However, in our study, commercialisation does not belong to the question of mission drift. It is more a question of market strategy. The MFI has the opportunity to stay with its market or move to another, whether commercially oriented or not. This is illustrated in the simple BCG matrix in Table 3. Table 3 states that mission drift refers to a departure from a market strategy based on serving poor clients with small loans. Staying with the same customers but expanding the financial services to include, say, insurance does not represent mission drift. The mission is defined relative to the segment. For the MFI, such a mission expansion may provide economies of scope, that is, a cheaper provision of services jointly (e.g., loans and insurance) than separately. Furthermore, staying with the segment does not imply that the MFI stays with the same customer. If the client builds a successful business, the

219

Myths in Microfinance

Table 3.

Categories of Mission Drift Given Market Segment and Product Range.

Products

Market Old

Old

No mission drift

New

Mission expansion: Offering a wider spectrum of financial services

New Mission drift: Moving into new customer segments Mission drift: Moving into new segments with new products

financial needs may require financial services that the MFI is unable to provide. This may lead to the graduation of clients to other financial institutions, while at the same time the MFI accepts new customers with demands for small loans. How can we measure mission drift? We follow Cull et al. (2007), who define mission drift as an average loan increase. This means that the MFI moves into other customer segments, either because it starts including wealthier customers, or because the existing clients become wealthier and therefore less likely to be part of the MFI’s target group. Other measures for mission drift also exist. For example, an MFI decreasing its outreach to women may be seen by some as mission drift. However, in this chapter we concentrate on the average loan. Is there a natural pattern in MFIs to gradually move towards wealthier clients? Fig. 1 provides an illustration on how average and mean loans vary with MFI experience. Fig. 1 shows no average loan increase. Mission drift would be indicated by an increase in average loan relative to the microfinance starting year. However, the figure shows that average loan size does not increase with the time that the MFI has been in business. Rather, the trend is downwards over the years. Thus, the time dimension of the average loan size does not support the supposition that the MFIs, on average, experience mission drift. Measuring drift along other time dimensions like the difference between the year of rating and the previous three years gives the same result. How can we explain this non-drift? In the following, we attempt to find arguments for the average loan increase in MFIs. From the bank’s production function (Freixas & Rochet, 1999), the higher the average loan, the higher the average profits and average costs. To this basic relationship

220

Fig. 1.

ROY MERSLAND AND R. ØYSTEIN STRØM

The Mean and Median of Average Loan (USD) Distributed by MFI Experience.

we add the MFI’s risk and lending methodology, and furthermore, its economies of scope, effort, incentives, as well as the external impact of competition and bank regulation. Positive economies of scope implies that adding new products decreases overall operating costs. If more products reduce costs, the MFI should also be able to cover costs for smaller loans, and so lower average loans should follow. We measure the effect based on the number of loan products and bank regulation, which may pick up economies of scope in lending and deposits. The MFI’s marketing efforts may bring the average loan down. The fraction of credit officers among employees is our signal of an MFI’s greater effort. Furthermore, if the credit officers receive pay linked to their performance in recruiting new customers, this can also bring the average loan down. In both cases, it is likely that credit officers will seek out new potential clients in the MFI’s existing customer segment. Since the first microfinance customers are likely to have the highest creditworthiness, the customers that follow should be eligible for lower average loans. Thus, as the MFI extends its business, a lower average loan could well be the result. Higher effort and better incentives could therefore mean that the MFI drifts towards poorer customer segments, that is, the opposite of what critics fear.

Myths in Microfinance

221

Can increased bank competition lead to less financing going to the poorer segments? Petersen and Rajan (1995) show that bank concentration is beneficial for marginal clients since the bank expects to recoup the early subsidisation with higher rates later. When competition increases, the bank is unable to uphold the expected long-term relationship on the same terms. Therefore, in a competitive market, early subsidisation is eliminated, harming the least well off and leading to higher average loans. Petersen and Rajan (1995) find support for their theory in tests of smaller firms. If a similar situation occurs in microfinance, we expect to see a contraction of loans to the poorest customers, inducing higher average loans. We test the hypotheses with the same statistical method as in Myth 2, that is, the random effects method. The results in Table 4 largely confirm the hypotheses. The (absolute) size of profits per customer increases the (absolute) size of the average loan. This is not surprising since more profits are likely to be earned on a loan of 100 than on a loan of 10. Table 4 further shows that average loan increases with costs per loan client, confirming the hypothesis. Thus, inefficient MFIs need to shift their loan portfolio towards a higher fraction of larger average loans. This means that inefficient MFIs are most susceptible to mission drift. Another implication is that higher cost efficiency makes MFIs better able to advance loans to the poorer parts of the community. Increasing the MFI’s scale of operation seems to be key here. This result is in line with Hishigsuren’s (2007) observations that the MFI in her case study increased lending to the poor, even when the time that the lending officer spent with the borrowing group was cut in half. Furthermore, the higher the risk, the lower the average loan. Thus, MFIs take on more risk when they lower the average loan. Lending to individuals tends, on the other hand, to increase the average loan, probably because the MFI is better able to differentiate between good and bad risks when lending to individuals. These results are as expected. The remaining explanations cannot be confirmed, although the signs are as expected. Nevertheless, the results show that increased scope does not induce higher average loans. Thus, the MFI may add new products or accept savings without risking an exodus from the microfinance market. Effort and incentives are not significant, but nearly so. Thus, pay for performance may be interpreted to imply lower average loans. It seems as if the credit officer looks for lending opportunities in the chosen customer segment more intensively with performance pay. If this is true, performancelinked pay brings microfinance to even poorer clients. Furthermore, the higher the fraction of credit officers, the lower the average loan. Thus, when

222

ROY MERSLAND AND R. ØYSTEIN STRØM

Table 4.

The Expanded Model of Average Loan. Scope

Effort

Incentives

Competition

All Variables

0.019 1.0411 4.2861 0.20110 0.005 0.000

0.053 0.9971 4.0191 0.22010

0.107 1.0181 4.0311 0.162

0.059 0.9991 4.0321 0.21410

Constant Average profit Average cost PaR30 Loan products Voluntary savings Credit officer fraction Performance pay Competition Individual loan Shareholder owned MFI age Assets DM Latin America DM Southern Africa DM MENA DM EECA

0.1931 0.113 0.002 0.0001 0.033 0.025 0.211 0.050

0.2081 0.122 0.003 0.0001 0.043 0.045 0.235 0.065

0.2021 0.119 0.002 0.0001 0.045 0.027 0.2587 0.068

0.018 0.1991 0.116 0.003 0.0001 0.015 0.042 0.220 0.076

0.018 1.0421 4.2851 0.21410 0.001 0.000 0.019 0.108 0.014 0.1991 0.098 0.003 0.0001 0.032 0.027 0.156 0.063

Wald (F) test sign. $N$ (Firm years)

0.000 813

0.000 794

0.000 852

0.000 816

0.000 742

0.036 0.096

Note: Is average loan in rated MFIs related to profit function variables, the number of loan products, effort, incentive, and competition in the MFI market? Regressions include risk and control variables using random effects, panel data estimation and the 3SLS methodology. Data are from 1998 to 2007. A raised number indicates a significant result, with the number representing the significance level, e.g., the superscript 5 means significant result at the 5% level. Loan products are the number of loan products the MFI offers; credit officer fraction is the fraction of credit officers among all employed; performance pay is a dummy variable where 1 indicates the credit officer is paid proportional to performance; competition is a self-constructed measure of local MFI competition scaled from 1 to 7 based on raters’ reports. The Wald F test is an exclusion test of the hypothesis that all coefficients together are equal to zero (Greene, 2003, p. 107). A low significance value rejects the hypothesis.

the MFI intensifies its marketing efforts, it tends to reach out to even poorer segments of the population than before. Efforts and incentives to provide loans mean that microfinance credit reaches out to deeper layers of the poor customer segment. In a further, unreported robustness test we leave out control variables, and find that the results largely coincide with those from the case when all variables are included. The regressions in Table 4 do not tell us if the MFI’s profitability increases when lending to wealthier customers. The MFI’s overall

Myths in Microfinance

223

profitability may be larger if it lends to 100 customers at 10 each than to 10 customers at 100 each. To explore the overall profitability of average loans, we run a regression yielding the following main results: ROA ¼ 0:41ð9:67ÞIM  0:51ð10:04Þ OPEX=A  0:10ð2:27ÞAL  0:10ð2:73ÞPaR30 þ Controls where ROA is the return on assets, IM the intermediation margin, that is, the interest differential between portfolio yield and the MFI’s cost of funds multiplied by loans to assets, OPEX/A the operating expenses to assets, and AL the average loan. The numbers in the parentheses are t-statistics, which show that the reported variables are significant. The variables are normalised to have an average of zero and a standard deviation of 1.0. Thus, the coefficients may be compared. We find that the MFI’s margins and costs first and foremost influence its overall profitability. Average loan plays a much lesser but still significant role, but the sign is opposite to the hypothesis. Overall profitability is reduced when average loan is increased, which means that the overall profitability of the MFI is reduced when it leaves its customer segment. Thus, a motive for higher profits cannot explain mission drift. On the contrary, the negative sign indicates that the MFI has sound economic reasons for staying in its client segment. This result may seem puzzling. Intuitively one would expect that a larger overall profit can be earned from issuing larger average loans. However, observers of microfinance have noticed that there is an undersupply of services to the poor (Helms, 2006; C-GAP, 2004, 2006). An undersupply translates into high portfolio yields, which among our firms is an average close to 40%. If the risk is also low, and the average PaR30 is indeed low, microfinance is a good business proposition. Lately it has also attracted ordinary banking firms. Thus, the main conclusion is that we cannot find evidence of mission drift in the descriptive or econometric evidence. These results are in line with the scant rigorous literature on mission drift (Cull et al., 2007; Hishigsuren, 2007). Thus, a profit-conscious MFI should not drift from the microfinance field, but instead stay. Furthermore, the more cost-effective the MFI, the lower the average loan. Thus, a prediction is that further efforts to reduce costs should result in even poorer segments. Rather than drifting away from the poor customer segment, it seems that microfinance is developing into a viable business model.

224

ROY MERSLAND AND R. ØYSTEIN STRØM

DISCUSSION AND CONCLUSION How can we understand the main results offered in this chapter? To summarise, MFIs do not need to be shareholder owned in order to compete successfully, horizontal governance relationships are important on par with the vertical, and MFIs do not drift away from their mission. Our interpretation is that microfinance is establishing itself as a viable part of the banking industry, and this development follows the experience in many developed countries. Microfinance can be seen as part of a development of the market economy. Furthermore, the diversity of organisational forms in the microfinance industry is not transient, but answers the need for customers to trust the MFI. Let us take a brief look at the historical evidence. Hansmann (1996) shows that in the USA, early financial institutions targeting the average customer were often savings and credit unions, that is, institutions built by customers themselves. From a historical perspective, Cull et al. (2006) show that throughout North Atlantic countries, different types of intermediaries emerged to supply finance for small businesses and individuals. In France, notaries played this role, while in Anglo-American countries the role was filled by small commercial banks as well as building societies and credit unions, and in Germany and Scandinavian countries the function was filled by savings banks. Furthermore, Rasmussen (1988) reports historical evidence that mutual banks attract smaller customers and take on less risk than stock banks when regulation is weak. Historical evidence also confirms the importance of horizontal links to customers, underscoring that MFIs have the benefit of filtering good customers from weak and inducing loan clients to repay (Hansmann, 1996; Desrochers & Fischer, 2002) since they are better able to tap into local information networks. In this chapter we have explored three myths about microfinance. With the arrival of large, global datasets on rated MFIs over several years, it is possible to apply statistical techniques to discover if the myths are true. The first myth is that MFIs need to be shareholder owned. Based on a number of performance measures such as return on assets, return on equity, cost measures, as well as measures of risk, loan methodology, and main market served, we cannot differentiate between shareholder-owned MFIs and member-owned firms. The second is that MFIs need to install Western-type governance mechanisms in order to control the potentially conflict-ridden vertical owner–management relationship. We add the bank–customer horizontal

225

Myths in Microfinance

relationship, operationalised as a female CEO and individual loans. With a short banking history for many MFI customers and the MFIs themselves, we believe the bank’s relationship with its customers is potentially important. The results support this belief, since we find the horizontal relationship to be important, together with the vertical. The third myth is that MFIs leave their mission of serving poor customers with small loans. We cannot find evidence for this. On the contrary, the average loan in the microfinance industry is constant or reduced over time, and this is partly driven by increased effort to lower the cost per customer. We believe that dispelling these myths should encourage a rethink of policies towards MFIs. The results point towards an industry finding its own viable business model, establishing its own way of providing banking services to a customer segment formerly unattached to the economy at large. The microfinance industry is showing that it is able to reach poor customers in a financially sustainable way.

REFERENCES Adams, R. B., & Mehran, H. (2003). Is corporate governance different for bank holding companies? Federal Reserve Bank of New York Economic Policy Review, 9, 123–142. Altunbas, Y., Evans, L., & Molyneux, P. (2001). Bank ownership and efficiency. Journal of Money, Credit, and Banking, 33, 926–954. Armendariz de Aghion, B., & Morduch, J. (2005). The economics of microfinance. Cambridge: MIT Press. Berger, A. N., & Humphrey, D. B. (1997). Efficiency of financial institutions: International survey and directions for future research. European Journal of Operational Research, 98(2), 175–212. Brickley, J. A., Coles, J. L., & Jarrell, G. (1997). Leadership structure: Separating the CEO and chairman of the board. Journal of Corporate Finance, 3(3), 189–220. C-GAP. (2004). Key principles of microfinance. C-GAP, Washington. C-GAP. (2006). Good practice guidelines for funders of microfinance. C-GAP, Washington. Christen, R. P., & Drake, D. (2002). Commercialization: The new reality of microfinance. In: D. Drake & E. Rhyne (Eds), The commercialization of microfinance: Balancing business and development (pp. 2–22). Bloomfield: Kumarian Press. Christen, R. P., Rosenberg, R., & Jayadeva, V. (2004). Financial institutions with a ‘‘Double Bottom Line’’: Implications for the future of microfinance. Washington: C-GAP. Crespi, R., Garcia-Cestona, M. A., & Salas, V. (2004). Governance mechanisms in Spanish banks. Does ownership matter? Journal of Banking and Finance, 28, 2311–2330. Cull, R., Davis, L. E., Lamoreaux, N. R., & Rosenthal, J.-L. (2006). Historical financing of small- and medium-size enterprises. Journal of Banking and Finance, 30, 3017–3042. Cull, R., Demigu¨c-Kunt, A., & Morduch, J. (2007). Financial performance and outreach: A global analysis of leading microbanks. The Economic Journal, 117, 107–133.

226

ROY MERSLAND AND R. ØYSTEIN STRØM

Desrochers, M., & Fischer, K. P. (2002). Corporate governance and depository institutions failure: The case of an emerging market economy. Quebec, Canada: CIRPEE. Dichter, T. W. (1996). Questioning the future of NGOs in microfinance. Journal of International Development, 8, 259–269. Dichter, T. W., & Harper, M. (Eds). (2007). What’s wrong with microfinance? Essex, England: Practical Action Publishing. Eisenberg, T., Sundgren, S., & Wells, M. T. (1998). Larger board size and decreasing firm value in small firms. Journal of Financial Economics, 48, 35–54. ESBG. (2004). European banking consolidation: A considered analysis of the retail banking market and the position of the European Savings Banks Group. Brussels: ESBG. Freixas, X., & Rochet, J.-C. (1999). The microeconomics of banking. Cambridge, Mass: The MIT Press. Fernando, N. A. (2004). Micro success story, transformation of nongovernmental organizations into regulated financial institutions. Manila: Asian Development Bank. Greene, W. H. (2003). Econometric analysis (5th ed.). New York: Prentice Hall. Hansmann, H. (1996). The ownership of enterprise. Cambridge, MA: Belknap Press, Harvard University. Hart, O. (1983). The market mechanism as an incentive scheme. Bell Journal of Economics, 14, 366–382. Hartarska, V. (2005). Governance and performance of microfinance institutions in central and Eastern Europe and the newly independent states. World Development, 33, 1627–1648. Hartarska, V., & Nadolnyak, D. (2007). Do regulated microfinance institutions achieve better sustainability and outreach? Cross-country evidence. Applied Economics, 39, 1–16. Helms, B. (2006). Access for all. Washington, DC: CGAP. Hishigsuren, G. (2006). Transformation of micro-finance operations from NGO to regulated MFI. Decatur, Georgia, USA: Ideas. Hishigsuren, G. (2007). Evaluating mission drift in microfinance: Lessons for programs with social mission. Evaluation Review, 31(3), 203–260. Iannotta, G., Nocera, G., & Sironi, A. (2007). Ownership structure, risk and performance in the European banking industry. Journal of Banking and Finance, 31, 2127–2149. Jensen, M. C. W., & Meckling, W. (1976). Theory of the firm: Managerial behavior, agency costs, and ownership structure. Journal of Financial Economics, 3, 305–360. Ledgerwood, J., & White, V. (2006). Transforming microfinance institutions. Washington, DC: The World Bank and The MicroFinance Network. Mayers, D., & Smith, C. W. (1983). Ownership structures across lines of property-casualty insurance. Journal of Law and Economics, 31, 351–378. Mersland, R. (forthcoming). The cost of ownership in microfinance organizations. World Development. Mersland, R., & Strøm, R.Ø. (2008a). Performance and trade-offs in microfinance organizations – Does ownership matter? Journal of International Development, 20, 598–612. Mersland, R., & Strøm, R. Ø. (2008b). Performance and corporate governance in microfinance institutions. MPRA Paper 3888, University Library of Munich. Mersland, R., & Strøm, R. Ø. (2008c). Microfinance mission drift? MPRA Paper, University Library of Munich. Morduch, J. (2000). The microfinance schism. World Development, 28, 617–629. OECD. (2004). Principles of corporate governance. Paris: OECD.

Myths in Microfinance

227

Otero, M., & Chu, M. (2002). Governance and ownership of microfinance institutions. In: D. Drake & E. Rhyne (Eds), The commercialization of microfinance. Bloomfield: Kumarian Press. Oxelheim, L., & Randøy, T. (2003). The impact of foreign board membership on firm value. Journal of Banking and Finance, 27(12), 2369–2392. Petersen, M. A., & Rajan, R. G. (1995). The effect of credit market competition on lending relationships. Quarterly Journal of Economics, 110, 407–443. Pischke, J. D. V. (1996). Measuring the trade-off between outreach and sustainability of microenterprise lenders. Journal of International Development, 8, 225–239. Rasmussen, E. (1988). Mutual banks and stock banks. Journal of Law and Economics, 31(2), 395–421. Robinson, M. S. (2001). The microfinance revolution: Sustainable finance for the poor. New York: World Bank. Rock, R., Otero, M., & Saltzman, S. (1998). Principles and practices of microfinance governance. Microenterprise Best Practices, DAI/USAID. Rhyne, E. (1998). The yin and yang of microfinance: Reaching the poor and sustainability. Microbanking Bulletin 2. Schmidt, K. M. (1997). Managerial incentives and product market competition. Review of Economic Studies, 64, 191–213. Schreiner, M. (2002). Aspects of outreach: A framework for discussion of the social benefits of microfinance. Journal of International Development, 14, 591–603. Solnik, B., & McLeavey, D. (2004). International investments (5th ed.). Pearson Education. Steinwand, D. (2000). A risk management framework for microfinance institutions. Eschborn, Germany: GTZ, Financial Systems Development. Tirole, J. (2006). The theory of corporate governance. Princeton: Princeton University Press. White, V., & Campion, A. (2002). Transformation – Journey from NGO to regulated MFI. In: D. Drake & E. Rhyne (Eds), The commercialization of microfinance. Bloomfield: Kumarian Press. Woolridge, J. M. (2002). Econometric analysis of cross section and panel data. Cambridge, MA: The MIT Press. Woller, G. (2002). The promise and peril of microfinance commercialization. Small Enterprise Development, 13(4), 12–21. Woller, G. M., Christopher, D., & Warner, W. (1999). Where to microfinance? International Journal of Economic Development, 1(1), 29–64. Yermack, D. (1996). Higher market valuation of companies with a small board of directors. Journal of Financial Economics, 40, 185–212.

This page intentionally left blank

PART IV TECHNOLOGY AND INTERNATIONAL EXPANSION

This page intentionally left blank

R&D AND FOREIGN DIRECT INVESTMENT WITH ASYMMETRIES IN KNOWLEDGE TRANSMISSION Maria Luisa Petit, Francesca Sanna-Randaccio and Roberta Sestini ABSTRACT The purpose of the chapter is to analyze how firms’ R&D investment decisions are affected by asymmetries in knowledge transmission, taking into account different sources of asymmetry, such as unequal know-how management capabilities and spillovers localization within an international oligopoly. We follow a game theoretic approach and consider a twocountry imperfect competition model with two firms – one from each country – producing a homogeneous good. Both the firms’ mode of foreign expansion and R&D level are endogenously determined. We find that a better ability to manage knowledge flows incentivates the firm to invest more in R&D. By introducing geographically bounded spillovers, we also show that one-way foreign direct investment (FDI) stimulates the multinational enterprise (MNE) to raise its own R&D, due to both the elimination of transport cost and a greater ability to source. Furthermore, it emerges that when geographical proximity increases the MNE’s New Perspectives in International Business Research Progress in International Business Research, Volume 3, 231–261 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1745-8862/doi:10.1016/S1745-8862(08)03011-2

231

232

MARIA LUISA PETIT ET AL.

capability to source local know-how, FDI is more likely to occur. The originality of this chapter relies on the analysis of the impact of asymmetries within an oligopoly model with endogenous R&D. Differently from other studies, this framework allows us to provide neat analytical results.

1. INTRODUCTION Over the past two decades a wide body of theoretical and empirical literature has stressed the effects of knowledge flows (or spillovers) on firms’ R&D activity (see, for comprehensive surveys, De Bondt, 1997; Castellani & Zanfei, 2006). This literature suggests that higher spillovers may negatively affect the propensity of firms to undertake in-house R&D activities (the socalled own R&D), through a disincentive effect. On the other hand, higher spillovers lead to a cost reduction resulting in a market enhancement effect, representing an incentive for R&D investment. With a few exceptions (De Bondt & Henriques, 1995) these studies are carried out under the assumption that spillovers are symmetric, that is the extent to which firms may source knowledge from each other is similar. In this chapter we aim at relaxing the assumption of symmetric spillovers, considering two situations in which firms may be led to manage external information flows by deliberately increasing incoming spillovers (and/or decreasing outgoing spillovers):  First, we allow for different know-how management capabilities.  Secondly, we consider the effects of geographical distance as a source of asymmetry, thus exploring the effects of geographically bounded spillovers in an international oligopoly. As to the first point, the hypothesis that firms generate and receive spillovers to the same extent (that is the symmetry between incoming and outgoing spillovers) obviously precludes the notion that firms are able somehow to manage these information flows. This view has been questioned in the recent literature. Growing empirical evidence indicates that firms exert effort and employ resources in R&D also in order to increase their ability to appropriate the knowledge and technology elaborated by other firms. This behaviour was first pointed out by Cohen and Levinthal (1989), when they argued that external knowledge flows are more effective for the innovation process of a firm when the firm itself engages in R&D activities. In other words, firms aim at increasing the amount of incoming spillovers by

R&D and FDI with Asymmetries in Knowledge Transmission

233

investing in ‘‘absorptive capacity’’. In such a scenario it is reasonable that the above predictions as to the disincentive effect of spillovers do not hold true any longer. This is because the desire to assimilate external knowledge flows creates a positive incentive to invest in R&D. Hence, introducing absorptive capacity may lead to an increase in equilibrium industry R&D investment due to the presence of spillovers. Moreover, a firm may attempt at increasing its incoming spillovers by voluntarily trading technological knowledge with partners, as it is typical in the research joint venture information-sharing cartels (Kamien, Muller, & Zang, 1992). An empirical study by Cassiman and Veugelers (2002) documents that incoming spillovers have a positive and significant effect on the probability of firms cooperating in R&D. In a further study, Veugelers (1997) explores in a sample of Flemish companies the closely related issue of the relationship between external sourcing and internal (or in-house) R&D activity, finding that only when one explicitly takes into account absorptive capacity it is possible that cooperative R&D engagements have a significant effect on internal R&D expenditures. This result gives support to the hypothesis of complementarity between in-house R&D and external know-how. In the theoretical research, however, the effects of absorptive capacity on the incentive to invest in R&D are far from being thoroughly assessed and clear-cut. While Cohen and Levinthal (1989) stressed the positive effect of absorptive capacity on the incentive to invest, Grunfeld (2003) reached the conclusion that this result does not always hold true, as two opposite forces act on the incentive to invest in R&D. On one side the R&D effort is spurred by the necessity to learn from others, while on the other side higher absorptive capacity leads the rivals to invest less. By incorporating absorptive capacity (through an endogenous spillover rate which is function of own R&D) into the d’Aspremont and Jacquemin (1988) model, Grunfeld shows that if the market size is sufficiently large, the negative traditional effect of spillovers outweighs the positive learning effect. As to the second rationale for our investigation, both empirical and theoretical studies have suggested that the distance between the receiver and the generator of knowledge may play an important role. The geographical dimension of know-how transfers is particularly relevant in the analysis of firms’ strategies of international expansion. When geographical proximity increases involuntary know-how transfer, one of the reasons that may push firms towards multinational expansion is the possibility to locate subsidiaries near sources of technological innovation. At the same time, domestic firms may take advantage of the closer location of these subsidiaries and absorb more easily and quickly the technological knowledge produced by them.

234

MARIA LUISA PETIT ET AL.

Also a wide body of empirical literature has shed light on the issue of international spillovers, finding that they tend to be geographically bounded, although the degree of technological localization has sharply fallen over time (see Jaffe, Trajtenberg, & Henderson, 1993; Keller, 2002, 2004). Similar results on the limited geographical scope of spillovers are obtained by studies which use patent citation information to trace the presence of spillovers from foreign subsidiaries to the local economy (for instance, Almeida, 1996). Moreover, the localization of spillovers is shown by patent citation studies which document that there are also technology transfers from local sources to foreign subsidiaries (Almeida, 1996; Branstetter, 2000; Frost, 1998) and by studies that have found that firms may invest abroad with the aim to absorb technological knowledge (Neven & Siotis, 1996; Frost, 1998). These studies indicate that knowledge flows between the multinational enterprise (MNE’s) subsidiary and the local producers take place in both directions. Cassiman and Veugelers (2002, 2004) show that foreign subsidiaries are more likely to be acquiring local know-how than to be transferring know-how to the local economy. This asymmetry in the intensity of external knowledge flows between foreign subsidiaries and local producers is confirmed by Singh (2007). This evidence suggests that technological spillovers can be modelled as dependent on the mode chosen by the firm to serve the foreign market, since a closer location increases the degree of knowledge transmission. Thus, we need to introduce a ‘‘proximity’’ effect into the model (i.e., we can assume that multinational companies – MNEs – and exporters operate with different degrees of technological spillovers) and an asymmetry between the incoming and the outgoing spillover for an MNE. The geographical dimension of spillovers has been accounted for only by a few theoretical models, which however consider R&D as exogenous and thus have a short run nature. In fact the papers by Ethier and Markusen (1996), Fosfuri and Motta (1999) and Siotis (1999) examine how localized spillovers may affect the firm’s decisions on how to serve a foreign market. Furthermore, in these models only one firm is allowed to expand abroad. The only analytical study addressing the impact of spillovers localization within a model allowing for endogenous R&D is Petit, Sanna-Randaccio, and Sestini (2005).1 Such study presents a dynamic model, which highlights the differences between short- and long-run effects. However, the complexity of the model does not allow for analytical results and thus the findings are only based on numerical simulations. We build here a static model, endogenizing both the level of R&D investment and the firms’ mode of foreign expansion, which allows us to

R&D and FDI with Asymmetries in Knowledge Transmission

235

obtain analytical results that more fully highlight the economic mechanisms through which asymmetric spillovers influence firms’ internationalization and innovation strategies. We start with a simplified version of the model where market structure is exogenous, focusing on the effects of asymmetric spillovers on the equilibrium R&D investment in oligopolistic markets. Then we present an extended version of the model where both the firms’ mode of foreign expansion and R&D level are endogenously determined in the presence of localized spillovers. We consider only the case of low spillovers as this appears to be the most relevant empirical case.2 Along the lines of Petit and Sanna-Randaccio (2000),3 but allowing for asymmetric spillovers, we set up a two-country imperfect competition model with two firms – one from each country – producing a homogeneous good. In particular, we specifically investigate process-enhancing or cost-reducing R&D investment. The extended model is structured as a three-stage game in which each firm must take three different types of decisions: (i) the mode of foreign expansion [export or foreign direct investment (FDI)], (ii) how much to invest in R&D and (iii) how much to sell in each market. The equilibrium market structure is therefore endogenously determined as the solution of the three-stage game. The chapter is organized as follows. Section 2 presents the first simplified version of the model. Section 3 introduces the extended model examining the effects of localized technological spillovers on R&D activities and on the optimal internationalization strategies of firms. Section 4 presents the main conclusions.

2. THE SIMPLIFIED MODEL: INCREASING INCOMING SPILLOVERS VIA KNOW-HOW MANAGEMENT We build here the simplest model that may capture the effects of asymmetric knowledge flows. There are two firms (iA{1,2}), which produce a homogeneous good in two identical countries (country I and II) and compete in quantities a` la Cournot. Learning resulting from investment in R&D characterizes the production process, implying that marginal and unit costs decrease as investment in R&D increases. That is, we consider process innovations that result in reductions in production costs. Notice that in our set-up the rivalry as to the innovation activity is of a non-tournament kind. This implies that there exist many different research

236

MARIA LUISA PETIT ET AL.

paths that firms can follow, each leading to an equivalent amount of R&D expenditure and generating thus an equivalent reduction in production costs. Hence, rivals cannot prevent any firm from obtaining equivalent improvements through spending equivalent amounts of resources in R&D (De Bondt, 1997). Let Ii be the investment in research undertaken by firm i (own R&D) and let mi (Ii) denote firm i’s marginal (unit variable) cost (i ¼ 1,2). The function mi (Ii) represents the (negative) relationship between firm i’s marginal cost and its level of R&D investment given by Ii. In addition, we allow for the possibility of imperfect appropriability (i.e., technological spillovers between the firms), introducing a spillover parameter aiA[0,1], i ¼ 1,2. In our framework, this parameter, though exogenously given, is meant to reflect the notion of different abilities of the firms to manage external information flows. Given this assumption on spillovers, the magnitude of firm i’s cost reduction is determined by its own R&D and by a fraction ai of the other firm’s investment. In our set-up then the parameter ai is an incoming spillover. More specifically, denoting with I ¼ ðI 1 ; I 2 Þ mi ðIÞ ¼ Ai  yðI i þ ai I j Þ i; j ¼ 1; 2; iaj

(1)

The parameter Ai in (1) can be considered as the initial marginal cost of firm i, i.e., the cost that would prevail with no investment in R&D.4 The parameter y, with yZ0, determines the rate at which mi declines with an increase in the R&D level. It represents the productivity of the firm’s research effort. The expression ðI i þ ai I j Þ represents firm’s i total knowledge, also called effective R&D. The inverse demand function for each market is linear and is defined as pI ¼ a  bðq1;I þ q2;I Þ

pII ¼ a  bðq1;II þ q2;II Þ

(2)

where pI and pII denote prices in country I and II, respectively, and qi,k represents the sales of firm i in country k (i ¼ 1, 2, k ¼ I, II). The parameters a and b are positive constants and measure the size of the market in each country.5 We assume parameter values which guarantee the non-negativity of prices and marginal costs and which ensure the possibility for both firms to be active.6 At this stage we assume that both firms are multinationals (i.e., each firm undertakes a direct investment – FDI – by creating a production subsidiary in the rival’s country). This implies that plant-specific fixed costs G should be included twice in the objective function of both firms.

R&D and FDI with Asymmetries in Knowledge Transmission

237

Each firm then maximizes profits, given by ¼ ða  bðq1;I þ q2;I ÞÞq1;I þ ða  bðq1;II þ q2;II ÞÞq1;II pDD 1  ðA1  yðI 1 þ a1 I 2 ÞÞðq1;I þ q1;II Þ 

gI 21  F  2G 2

ð3Þ

pDD ¼ ða  bðq1;I þ q2;I ÞÞq2;I þ ða  bðq1;II þ q2;II ÞÞq2;II 2  ðA2  yða2 I 1 þ I 2 ÞÞðq2;I þ q2;II Þ 

gI 22  F  2G 2

ð4Þ

where the superscript DD stands for MNE duopoly. Notice that in (3) and (4) we modelled the cost of investment in R&D as gIi2 =2, with gW0. This simple quadratic investment cost function guarantees decreasing returns to R&D expenditure (see, e.g., Cheng, 1984).7 We want to allow for the possibility of imperfect appropriability in the form of (asymmetric) technological spillovers between the firms. Obviously, the case of no spillovers (ai ¼ 0) may only arise in a situation of strong intellectual protection. More frequently, however, involuntary information leaks occur as a result of reverse engineering, industrial espionage or by hiring away employees of an innovative firm. The cases of partial to full spillovers can be modelled by setting 0oair1. Note that in this chapter, we consider only the case of aio0.5, as it represents the empirically most relevant scenario. Moreover, we recall that the asymmetry here is the outcome of different abilities by firms to absorb or assimilate intra-industry spillovers.8 Equilibrium will be determined by solving a two-stage Cournot duopoly game. In the first stage, firms decide how much to invest in R&D, knowing that these decisions are irreversible. In the second stage, each firm optimally chooses the amount of sales in both countries (and therefore of output). The subgame perfect equilibrium output and investment levels are obtained using backwards induction. For the sake of simplicity, we take into account the simplest case where both firms are identical but for different know-how management capabilities. This equals to saying that no firm has an initial cost advantage over the other, i.e. A1 ¼ A2.

2.1. Strategic Investments with Asymmetric Spillovers We begin by solving the second stage of the game, at which each firm i chooses the level of its sales at home and abroad, and thus its own level of

238

MARIA LUISA PETIT ET AL.

output, maximizing its objective function as in (3) and (4) under the Cournot assumption. We thus obtain q1;I ¼ q1;II ¼

a  A þ yð2  a2 ÞI 1 þ yð2a1  1ÞI 2 3b

(5)

q2;II ¼ q2;I ¼

a  A þ yð2  a1 ÞI 2 þ yð2a2  1ÞI 1 3b

(6)

Notice that for each firm sales increase with its own level of R&D. As a matter of fact, by increasing its own innovative effort the firm succeeds in decreasing unit costs and increasing consequently its competitiveness in the product market. This occurs for any value of the outgoing spillover, though the positive effect is weakened as the outgoing spillover becomes high. On the other hand, the value of the incoming spillover is crucial in order to establish the sign of the effect on sales of each firm stemming from the investment in R&D of the rival. In particular, if the incoming spillover for firm i is given by aio0.5 this effect is negative, whilst it is positive for aiW0.5. This may be explained by noting that, from the standpoint of firm i, when the spillover parameter is rather small (aio0.5), the negative effect on firm’s i sales stemming from an increase in the rival’s competitiveness (due to R&D investment) prevails over the positive effect due to the portion of knowledge absorbed from firm j’s innovative activities. The opposite holds, obviously, when the spillover parameter is greater than 0.5 and the positive effects prevails over the negative one. To obtain Nash equilibrium strategies for I1 and I2, we maximize objective functions as given by (3) and (4) with respect to R&D investments, having substituted for quantities q1,I, q1,II, q2,I and q2,II as stated here above. We thus obtain best response functions for investment in R&D Ii ¼

4ða  AÞð2  aj Þy 4ð2  aj Þð2ai  1Þy2 þ Ij 9bg  4ð2  aj Þ2 y2 9bg  4ð2  aj Þ2 y2

(7)

i, j ¼ 1,2 (i 6¼ j). Notice that the second order condition (SOC), guaranteeing the local concavity of the objective function is given by 9bg  4(2  ai)2 y2W0, i,j ¼ 1,2, i 6¼ j. Moreover, it is easy to establish that the slope of firm’s i reaction function is negative (positive) if aio0.5 (aiW0.5), independently

239

R&D and FDI with Asymmetries in Knowledge Transmission

from the value assumed by aj.9 From Eqs in (7) we get the equilibrium values of R&D levels 4ða  AÞð2  aj Þ½3bg  4y2 ð2  ai Þð1  ai Þy DD I^i ¼ C

i ¼ 1; 2

(8)

where C ¼ 27 b2g212 bg y2 [(2ai)2þ(2aj)2]þ16 y4(2ai)(2aj)(1aiaj). Due to stability conditions it is possible to establish that C40. Equilibrium solutions for output and prices are reported in Appendix A. Our aim is to focus on the consequences of asymmetric spillovers, i.e., ai 6¼ aj. The parameter ai is viewed here as the result of a firm-specific absorption factor, due to different know-how management capabilities. That is, one of the two firms is absorbing more knowledge from the rival (i.e., the intensity of its incoming spillover is higher, which implies that the intensity of its outgoing spillover is lower, as compared to the other firm). The effect of a rise in incoming spillover intensity on the firm’s innovative performance is illustrated in Fig. 1, where the R&D reaction functions of the two firms are represented. We are mainly concerned on whether a rise in

I2 I1DD(I2)

I2DD(I1)

I1

Fig. 1.

The Effect of a Rise in a1 on R&D Equilibria.

240

MARIA LUISA PETIT ET AL.

absorptive capacity drives up the incentive to invest in own R&D. We will consider the case of a rise in a1 due to an increased efficiency in know-how management, while a2 does not change. The model shows that firm 1 (2) reaction function shifts upwards (downwards) rotating around its horizontal intercept (see proof). It follows that if the firm enjoys a higher ability to source external knowledge is also stimulated to increases its own R&D, while the reverse is the case for the other firm. In other words, the asymmetry generated by the rise in a1 enhances firm’s 1 profitability and thus the stimulus to innovate. We are able to state Proposition 1. An asymmetry due to an increase in incoming spillover intensity – ceteris paribus – leads to higher own R&D. Proof of Proposition 1. Let us rewrite best response functions as in (8) in a simplified form ¼ I DD 1

M 1 N 1 DD þ I J1 J1 2

(9)

I DD ¼ 2

M 2 N 2 DD þ I J2 J2 1

(10)

and

where Mi ¼ 4y(2  aj)(a  4), Ji ¼ 9gb  4y2(2  aj)2, Ni ¼ 4y2 (2  aj)(2ai  1)

i, j ¼ 1,2, i 6¼ j.

Equilibrium investments are then given by MiJ j þ NiMj DD I^i ¼ JiJj  NiNj

i; j ¼ 1; 2; iaj

(11)

Notice that focusing on strategic substitutes R&D investments implies that Nio0. Examining the effect of an increase in the incoming spillover (i.e., a1) on own R&D equilibrium investment of firm 1, it is easy to find that, as a1 gets higher, firm 1’s reaction function shifts upwards, pivoting around its horizontal intercept as its slope decreases ð@ðjJ 1 =N 1 jÞ=@a1 ¼ ð8y2 ð2  a2 ÞJ 1 Þ=N 21 40; because of SOCÞ, and the vertical intercept

R&D and FDI with Asymmetries in Knowledge Transmission

241

increases ð@ðM 1 =N 1 Þ=@a1 ¼ ða  AÞ=ðyð1  2a1 ÞÞ40Þ. On the other hand, firm 2’s reaction function shifts downwards, pivoting around its horizontal intercept ð@ðjM 2 =N 2 jÞ=@a1 o0 along with @ðjN 2 =J 2 jÞ= @a1 o0Þ: & It is also possible to show that Proposition 2. A rise in outgoing spillover intensity – ceteris paribus – results in lower own R&D. Proof of Proposition 2. Straightforward, as this equals to examining how firm 2 reacts to an increase in a1. & To summarize, the equilibrium level of own R&D investment chosen by firm 1 raises after an increase in the level of the incoming spillover intensity, whilst higher outgoing spillovers intensity depresses the incentive to invest in R&D, due to the fear of dissipation of its own technological knowledge.

3. THE EXTENDED MODEL: INCREASING INCOMING SPILLOVERS BY LOCATING ABROAD In this section, we explore another source of asymmetry in the involuntary transmission of technological knowledge. Firms are supposed henceforth to be identical in all respects. However, technological spillovers still differ in intensity since they depend on the mode chosen by firms to serve the foreign market. Now firms face three types of decisions: how to expand abroad, how much to invest in R&D and finally how much to sell in each country. At the first stage of the game, the duopolists choose the mode of foreign expansion, with a strategy space made of two possible strategies: export (EXP), i.e., producing in the home country and exporting abroad, and FDI, i.e., producing in both countries, thus becoming a multinational. We also assume that the choice of exporting implies additional unit (and marginal) transport costs denoted by s, whilst, as explained in Section 2, the FDI choice brings about additional plant-specific fixed costs G. Beside the case of a MNE duopoly, where the objective functions are as specified in (3) and (4), we introduce now two further possible (different) market configurations, with their respective objective functions.

242

MARIA LUISA PETIT ET AL.

An exporting duopoly arises when both firms have only one plant in a country and export to the other country. Profits are then given by pEE 1 ¼ ða  bðq1;I þ q2;I ÞÞq1;I þ ða  bðq1;II þ q2;II ÞÞq1;II  ðA1  yðI 1 þ a1 I 2 ÞÞq1;I  ðA1  yðI 1 þ a1 I 2 Þ þ sÞq1;II 

gI 21 F G 2

ð12Þ

pEE 2 ¼ ða  bðq1;I þ q2;I ÞÞq2;I þ ða  bðq1;II þ q2;II ÞÞq2;II  ðA2  yða2 I 1 þ I 2 ÞÞq2;II  ðA2  yða2 I 1 þ I 2 Þ þ sÞq2;I 

gI 22 F G 2

ð13Þ

where the superscript EE stands for exporting duopoly. Lastly, we allow for the case of a mixed duopoly, with an MNE and a exporting firm. In this scenario, one firm serves the other country by creating a new plant and the rival firm by exporting. Assuming firm 1 to be the exporting firm and firm 2 the MNE (i.e., the ED duopoly10), profits are given by pED 1 ¼ ða  bðq1;I þ q2;I ÞÞq1;I þ ða  bðq1;II þ q2;II ÞÞq1;II  ðA1  yðI 1 þ a1 I 2 ÞÞq1;I  ðA1  yðI 1 þ a1 I 2 Þ þ sÞq1;II 

gI 21 F G 2

ð14Þ

pED 2 ¼ ða  bðq1;I þ q2;I ÞÞq2;I þ ða  bðq1;II þ q2;II ÞÞq2;II  ðA2  yða2 I 1 þ I 2 ÞÞðq2;I þ q2;II Þ 

gI 22  F  2G 2

ð15Þ

We leave out the analysis of the MNE duopoly, having already determined best response functions (see Eq. (7)), equilibrium investments in R&D (Eq. (8)) and equilibrium quantities (Eq. (A.1)) and prices (Eq. (A.2)). Exporting Duopoly. Under this market configuration, likewise as in Section 2, we first obtain the values of the sales variables as functions of I1 and I2 and substitute them into the profit functions. We then get the reaction functions for investment in R&D Ii ¼

4ða  AÞð2  aj Þy 2sð2  aj Þy 4ð2  aj Þð2ai  1Þy2  þ I j (16) 9bg  4ð2  aj Þ2 y2 9bg  4ð2  aj Þ2 y2 9bg  4ð2  aj Þ2 y2

243

R&D and FDI with Asymmetries in Knowledge Transmission

i, j ¼ 1,2 (i 6¼ j). Employing a more compact notation, we can rewrite Eq. (16) as I EE 1 ¼

M 1 T 1 N 1 EE  þ I J1 J1 J1 2

(17)

I EE 2 ¼

M 2 T 2 N 2 EE  þ I J2 J2 J2 1

(18)

and

where Mi, Ji, Ni are as defined above (see Eqs. (9)–(10)), and Ti ¼ 2s(2  aj)y i, j ¼ 1,2, i 6¼ j. Hence equilibrium R&D investments are given by 2yð2  aj Þ½3bg  4y2 ð2  ai Þð1  ai Þ½2ða  AÞ  s EE I^i ¼ C

i ¼ 1; 2

(19)

where C ¼ 27b2g212bg y2 [(2ai)2þ(2aj)2]þ16 y4 (2ai) (2aj) (1aiaj). Employing a more compact notation we get MiJ j þ NiMj T iJ j þ NiT j EE I^i ¼  JiJj  NiNj JiJj  NiNj

i ¼ 1; 2

(20)

EE Notice that a positive equilibrium exists for I^i iff [3bg  4y2(2  ai) (1  ai)]W0, which is the same condition ensuring positivity of equilibrium R&D investments and quantities in the MNE duopoly examined in the previous section.11 Mixed Duopoly. We shall consider now the case in which one firm expands abroad via FDI, while the other does so via exports. As already mentioned, we assume that firm 1 exports the output produced in its home country, while firm 2 becomes a MNE. Again, we solve the game backwards. In the third stage, each firm chooses its own levels of sales at home and abroad, maximizing its profit function under the Cournot assumption. We thus obtain the values of the sales variables as functions of I1 and I2. Substituting them into the profit functions and maximizing with respect to I1 and I2 we get the best response functions for investment in R&D, i.e.

I1 ¼

4yða  AÞð2  a2 Þ 4syð2  a2 Þ 4ð2  a2 Þð2a1  1Þy2  þ I2 9bg  4ð2  a2 Þ2 y2 9bg  4ð2  a2 Þ2 y2 9bg  4ð2  a2 Þ2 y2 (21)

244

MARIA LUISA PETIT ET AL.

and I2 ¼

4yða  AÞð2  a1 Þ 2syð2  a1 Þ 4ð2  a1 Þð2a2  1Þy2 þ þ I1 9bg  4ð2  a1 Þ2 y2 9bg  4ð2  a1 Þ2 y2 9bg  4ð2  a1 Þ2 y2 (22)

or, employing a more compact notation I ED 1 ¼

M 1 V 1 N 1 ED  þ I J1 J1 J1 2

(23)

I ED 2 ¼

M 2 T 2 N 2 ED  þ I J2 J2 J2 1

(24)

and

where Mi, Ji, Ni, Ti are as already defined and Vi ¼ (4s(2  aj)y), i, j ¼ 1,2, i 6¼ j. Under this market configuration, Nash equilibrium values for investment in R&D are given by 2yð2  a2 Þ½3bg  4y2 ð2  a1 Þð1  a1 Þ½2ða  AÞ  s ED I^1 ¼ C 2 2sð3bg  4y a1 ð2  a1 ÞÞ  C

(25)

4ða  AÞð2  a1 Þ½3bg  4y2 ð2  a2 Þð1  a2 Þy ED I^2 ¼ C 2sð2  a1 Þy3bg  4y2 a2 ð2  a2 Þ þ C

(26)

or, employing a more compact notation M1J 2 þ N1M2 V 1J 2 þ N 1T 2 ED  I^1 ¼ J 1J 2  N1N 2 J 1J 2  N1N2

(27)

M2J 1 þ N2M1 T 2J 1  N 2V 1 ED I^2 ¼  J 1J 2  N1N 2 J 1J 2  N1N2

(28)

and

Since we want to focus on the role of asymmetric knowledge flows in encouraging (or discouraging) innovative efforts, we proceed now with some

R&D and FDI with Asymmetries in Knowledge Transmission

245

relevant assumptions regarding the geographical dimension of know-how transfers, which can be reasonably regarded as a source of asymmetry.

3.1. Assumptions on Spillover Localization We assume henceforth that the effects of geographical distance on the transmission of information between the firms should be taken into account. As a matter of fact, in assessing which agent benefits more from the rivals’ knowledge stock, a notion of distance (or proximity) between the technology receiver and the generator has been shown to be relevant in many empirical studies (see, among others, Singh, 2007; Almeida, 1996; Neven & Siotis, 1996; Frost, 1998). Aiming at introducing this notion as a source of asymmetry in knowledge flows we make the following assumptions on the spillover parameter.12 A1. The transfer of know-how between two exporters is lower than between two MNEs, i.e., DD aEE ij oaij ;

i ¼ 1; 2; j ¼ I; II

where aij is the portion of the knowledge produced by other firm(s) which is absorbed by firm i in country j, representing thus an incoming spillover parameter for firm i in country j. Recalling that in the ED case firm 1 is the exporter and firm 2 the MNE, the following assumption also holds: A2. ED aED 1I oa2I

In country I, the degree of transmission of technology from the local firm (firm 1) to the MNE (firm 2) ðaED 2I Þ is stronger than the degree of transmission from the MNE to the local firm ðaED 1I Þ. This is due to the fact that firm 1 (the local firm) cannot fully exploit some of the know-how transfer mechanisms (like personnel mobility or industrial espionage), since the bulk of research activities undertaken by firm 2 is located in its home country. Moreover, subsidiaries may try to prevent know-how from leaking out (for instance reducing the mobility of personnel by paying higher wages, or applying other strategies to minimize spillovers). On the other hand, the MNE (firm 2) can take advantage of all the knowledge transfer mechanisms, since it produces in country I where the local firm (firm 1) has its main centre of research activity. Empirical research gives

246

MARIA LUISA PETIT ET AL.

support to this hypothesis, as shown in Cassiman and Veugelers (2004) and Singh (2007). A3. When both firms are multinationals (the DD case), they can take advantage of all the information transfer mechanisms since both firms have a plant in the rival’s home country. Therefore, the transfer of knowhow between two MNEs is the same as for the MNE in the mixed oligopoly (ED). DD DD aED 2I ¼ a2I ¼ a1II

A4. We also hypothesize that there is no cost of technology transfer from the parent firm to the subsidiary, and vice versa. Therefore, the fraction of knowledge that firm i receives in country I is the same as the fraction it receives in country II. It thus follows that ED aED 2I ¼ a2II DD aDD 1I ¼ a1II DD aDD 2I ¼ a2II

This assumption makes it possible to eliminate the country indexes.13 We can therefore simplify the relationships between the spillover parameters as follows: DD aEE i oai ;

i ¼ 1; 2

ED aED 1 oa2 DD aED ¼ aDD 2 ¼ a1 2

ED As to the relationship between aEE 1 and a1 , the following assumptions can be made: (i) the fraction of technological information received by the exporting firm 1 in the case of two exporters ðaEE 1 Þ is the same as that EE ED received by the exporting firm 1 when firm 2 is a MNE ðaED 1 Þ, i.e., a1 ¼ a1 , or (ii) since in the ED case the subsidiary of firm 2 is producing in country I, some more information leaks from firm 2 towards firm 1 may occur if

R&D and FDI with Asymmetries in Knowledge Transmission

247

ED compared with the EE case, i.e., aEE 1 oa1 . Therefore, it seems appropriate to assume ED aEE 1  a1

Therefore, taking into account all the above inequalities, we can write ED DD  aED aEE i 1 oa2 ¼ ai

Hence, even if the spillover parameters are not explicit functions of location, as in Duranton (2000), they are related to location (in our case, to the mode of foreign expansion) by the above defined constraints that ai and aj must satisfy. We also point out that, even if in the export–export case the distance between the two firms is the highest, this does not necessarily mean that there is no transmission of knowledge. The transmission can always take place through some of the usual channels of know-how transfer, i.e., reverse engineering (from imported goods), and also from international personnel mobility, journals and conferences. 3.2. Spillovers Localization and the Incentive to Innovate In this section, our assumptions on spillover localization will be inserted into the model, mainly investigating the impact that asymmetries in the degree of knowledge transmission – due to differences in location – may have on the incentive to innovate under different market configurations. First we carry out a comparison between firms’ innovative performance under the DD case and under the EE case. Denoting with aDD the (common) value of the parameter aDD ¼ aDD and with aEE the value of the parameter 1 2 EE EE a1 ¼ a2 , we substitute for ai and aj, i, j ¼ 1,2, i 6¼ j into Eqs. (8) and (19). We thus obtain DD DD I~1 ¼ I~2 ¼

4ða  AÞð2  aDD Þy 9bg  4ð2  aDD Þð1 þ aDD Þy2

(29)

and EE EE I~1 ¼ I~2 ¼

4ða  AÞð2  aEE Þy 2sð2  aEE Þy  9bg  4ð2  aEE Þð1 þ aEE Þy2 9bg  4ð2  aEE Þð1 þ aEE Þy2 (30)

248

MARIA LUISA PETIT ET AL.

We can thus state the following proposition: Proposition 3. In an international duopoly with localized spillovers, DD ~EE depends on the relative magnitude of the market whether I~i 4 o Ii expansion effect versus the free-riding effect. Proof of Proposition 3. Let us define  4ða  AÞð2  aDD Þy DD EE ~ ~ D ¼ Ii  Ii ¼ 9bg  4ð2  aDD Þð1 þ aDD Þy2  4ða  AÞð2  aEE Þy þ  9bg  4ð2  aEE Þð1 þ aEE Þy2 2sð2  aEE Þy þ 9bg  4ð2  aEE Þð1 þ aEE Þy2 Denoting with OK the ratio ð4ða  AÞð2  aK ÞyÞ=ð9bg  4ð2  aK Þ ð1 þ aK Þy2 Þ, with K 2 fDD; EEg, we have that D ¼ ½ODD  OEE þ ð2sð2  aEE ÞyÞ=ð9bg  4ð2  aEE Þð1 þ aEE Þy2 Þ. Reminding that our assumptions on spillover localization imply aEE oaDD , then the first term in D – the term between square brackets – representing the free-riding effect, is nonpositive, since @OK =@a ¼ ð4yða  AÞð4y2 ð2  aÞ2  9bgÞÞ=ð½9bg  4y2 ð2  aÞ ð1 þ aÞ2 Þ  0, because of SOC. The second term, which can be seen as the market enlargement effect of FDI, is obviously positive. & This proposition sheds light on two contrasting forces affecting the innovative performance of firms (MNEs versus exporters) in an international duopoly with localized spillovers. In particular, the free riding – or disincentive – effect stemming from spillovers implies that higher (both incoming and outgoing) spillovers drives down own R&D effort. This is the typical negative effect according to which non-cooperative strategic R&D levels typically decrease with the magnitude of the spillovers, since the presence of leakages tends to limit the appropriability of individual activities. Given our assumptions on geographically bounded spillovers, this effect displays greater negative consequences as proximity increases. On the other hand, the FDI choice – as compared to export – by eliminating transport costs removes the cost advantage enjoyed by the locally based producer vis-a`-vis the foreign one, thus increasing competition in the product market. This results in an increase in the foreign sales of MNEs as compared to exporters. As a matter of fact, the possibility to serve a larger market typically increases the profitability of the research

249

R&D and FDI with Asymmetries in Knowledge Transmission

expenditures and therefore may become an incentive for the MNEs to invest more in research than the exporting firms, due to the so-called ‘‘market expansion effect’’. Again under the hypothesis of spillover localization, equilibrium values for the sales of each firm at home and abroad and then equilibrium prices are computed (for each market configuration considered). Analytical results along with a comparison between outputs and prices under the EE case and the DD case, respectively, are reported in Appendix B. We now carry out a comparison in terms of innovative performance between the mixed duopoly and the case of a duopoly made of two exporters. As we are interested in disentangling the effect of an increase in incoming spillovers due to greater proximity to the rival’s R&D main labs, ED DD ¼ aED . we maintain now that aEE i 1 oa2 ¼ ai Moving from the graphical analysis of the best response functions in the EE case (see Eqs. (17) and (18)) and in the ED case (Eqs. (23) and (24)), Fig. 2 shows that the reaction curve of firm 2 in the ED duopoly is characterized by both a higher vertical intercept and (in absolute value) a lower slope. The first effect (on the intercept) captures the benefit for firm 2 due to transport costs elimination, when moving from export to FDI. The slope effect instead is due to firm’s 2 increased ability to source local

I2 ED

I1EE(I2)

EE I1ED(I2)

I2ED(I1) I2EE(I1)

I1

Fig. 2.

A Comparison of R&D Investments in the EE and in the ED Cases.

250

MARIA LUISA PETIT ET AL.

knowledge due to a rise in a2. At the same time, the intercept of firm’s 1 reaction curve in the ED case shifts downwards (as firm 1 is not any more protected in the home market by transport costs) and the slope (in absolute value) increases because of higher outgoing spillovers. We are then able to state that Proposition 4. When localized spillovers lead to an increase in the incoming spillover of the investor, one-way FDI results in a rise in the MNE own R&D and in a decrease in the local firm own R&D. Proof of Proposition 4. Examining the best response function of firm 1 in the EE case (Eq. (17)) and in the ED case (Eq. (23)), respectively, the vertical intercept in the latter market structure is lower than the vertical intercept in the former case, being ðM 1 =N 1 Þ þ ðT 1 =N 1 Þ4  ðM 1 =N 1 Þ þðV 1 =N 1 Þ, as both T 1 =N 1 and V1 =N 1 are negative. As to the slope, the expression J 1 =N 1 differs in the EE case and in the ED case, since ED aEE 2 oa2 . We easily find that ð@ðjJ 1 =N 1 jÞÞ=@a2 40. Then, considering the best response function of firm 2 under both market structures, we obtain that the vertical intercept in the ED case is higher than the vertical ED intercept in the EE case, having assumed that aEE 1 ¼ a1 . Likewise as previously seen, the slopes differ, with ð@ðjN 2 =J 2 jÞÞ=@a2 o0. As depicted in ED EE EE ED Fig. 2, it follows that I~2 4I~1 ¼ I~2 4I~1 . & A comparison between the ED and the DD case gives rise to ambiguous results, because of the presence of two contrasting forces of opposite sign – the market enlargement effect versus the traditional negative spillover effect. Therefore, the ranking of R&D investment equilibria strongly depends on the relative magnitude of these effects, likewise in the comparison between equilibrium R&D investments in the DD and in the EE case (see Proposition 3). We investigate finally the consequences of (increasing) spillover localization within a market structure – the mixed duopoly or ED case – where the ED degree of asymmetry may play a crucial role (as aED 1 oa2 ). In particular, we want to ascertain whether or not deepening the degree of localization could give rise to different effects on the firms’ innovative performance. Looking at best response functions as in Eqs. (23) and (24), we examine how an increase in the degree of localization affects R&D equilibrium investment of firms, given our assumptions on the mechanisms of knowledge transmission. In our model this amounts to analyzing the effects of an increase in aED 2 , ceteris paribus.

251

R&D and FDI with Asymmetries in Knowledge Transmission

I2 I1ED(I2)

I2ED(I1)

I1

Fig. 3.

The Effect of the Degree of Spillover Localization R&D Equilibria: ED Case.

As Fig. 3 depicts, the outcome of these changes is that the equilibrium R&D investment by the MNE – firm 2 – is prompted after an increase in the degree of spillovers localization, whilst the investment of the exporting firm is hampered. This may be explained as the result of the asymmetry favouring the multinational firm, due to greater possibilities to source the rival’s knowledge through FDI. We can thus state that Proposition 5. In a mixed duopoly, a rise in the degree of spillover localization, leading to a higher incoming spillover for the MNE, brings about a rise in own R&D of the MNE and to a fall in own R&D of the local firm. Proof of Proposition 5. After an increase in aED 2 , firm 1’s reaction function shifts downwards, pivoting around its vertical intercept, with an increasing slope (in absolute value). In fact, we have that ð@ðjJ 1 =N 1 jÞ=@a2 ¼ ð32y4 ð2  a2 Þð1  2a1 Þ þ 4y2 ð1  2a1 ÞJ 1 Þ=N 21 40, with J 1 40 because of SOC), and a decreasing horizontal intercept ð@ððM 1  V 1 Þ=J 1 ÞÞ=@a2 ¼ ðA  a  sÞð4yJ 1 þ 32y3 ð2  a2 Þ2 Þ=J 21 o0. On

252

MARIA LUISA PETIT ET AL.

the other hand, firm 2’s reaction function shifts upwards, pivoting around its vertical intercept (as @ðjN 2 =J 2 jÞ=@a2 o0 and @ðM 2 =N 2 jÞ= @a2 40). &

3.3. The Effect of Localized Spillovers on the Equilibrium Strategies for Foreign Expansion Since our aim is to endogenize the market structure, we shall examine now how the firms will perform their choices on the mode of foreign expansion and how our assumptions on geographically localized spillovers may affect these choices. To analyse firms’ international strategy decisions we need to calculate the profits of each firm corresponding to two different strategies, i.e., EXP (exporting) or FDI (direct investment). Then we will single out Nash equilibrium solution(s) of a matrix game between the two firms, where the pay-offs are the profits of each single firm and the strategy space is S i ¼ fEXP; FDIg; i ¼ 1; 2. A Nash equilibrium will determine a subgame perfect equilibrium market structure of the game under analysis. Due to the complexity of equilibrium profits we had to resort to numerical analysis. Being equilibrium profits a function of the value of the spillover parameter ai ; i ¼ 1; 2, we assigned to it different values depending on the mode of foreign expansion chosen by the firms and thus coherently with the assumptions explained in Section 3.1. In particular, the values of technological spillovers employed in our simulations are as reported in Table 1. In choosing the numerical values for this table we introduce the hypothesis that industry-specific features influence the amount of knowledge leaks (e.g., legal appropriation regime, complexity of know-how affecting the degree of appropriation). In particular, we assume that industry-specific features influence the lowest feasible value of the spillover parameter, i.e., aEE i ; i ¼ 1; 2. Therefore, we consider low, medium and high intensity of Table 1. Values Assigned to the Spillover Parameters ai and aj in the Simulations. Rows: industry-specific intensity of spillovers. Columns: degree of localization effect. Low Medium High

ED DD aEE ¼ 0:05; aED ¼ 0:2 i 1 ¼ 0:1; a2 ¼ ai EE ED ED DD ai ¼ 0:1; a1 ¼ 0:2; a2 ¼ a2 ¼ 0:3 ED DD aEE ¼ 0:2; aED ¼ 0:4 i 1 ¼ 0:3; a2 ¼ ai

R&D and FDI with Asymmetries in Knowledge Transmission

253

knowledge diffusion industries, according to the value assigned to this parameter, as indicated by the rows of Table 1. Moreover, the degree of the localization effect is allowed to vary, depending on the value assigned to the ¼ aED parameter aDD i 2 , as indicated in the columns in Table 1. This hypothesis is tailored to ascertain whether or not deepening the degree of localization at the intra-industry level (i.e., within the same sector) could give rise to different effects in terms of market structure. This issue is ¼ aED investigated by increasing the value of aDD i 2 . As to the other parameters, whenever possible they have been chosen on the basis of available empirical results.14 In the simulations reported here (see Tables 2–4), they have been assigned the following numerical values: a ¼ 36; b ¼ 2; A ¼ 5; s ¼ 2; g ¼ 1; y ¼ 0:3; G ¼ 15 and F ¼ 10. A large number of computations have been carried out in order to assess how different factors (such as host market size, plant scale economies, etc.) affect equilibrium outcomes. For lack of space we discuss here only a selection of these computations, which allows to ascertain how changes in some parameters (in particular the parameters reflecting the degree of spillovers asymmetry) influence the equilibrium market structure. Comparing the equilibria shown in Tables 2–4 with a scenario with nonlocalized spillovers (see Petit & Sanna-Randaccio, 2000), we have that, all the other parameters being equal, with geographically bounded spillovers a FDI–FDI equilibrium occurs while in the non-localized spillovers case we had an Export–Export solution.15 We argue that, in an environment characterized by asymmetric knowledge flows, there is an additional motivation for choosing the FDI strategy, represented by the possibility to absorb more technological knowledge from the rival firm. MNEs can thus make higher profits in relation to exporters as they may have access to a larger share of the research produced by the competitor. The Effect of Localized Spillovers on the Equilibrium Market Structure with Low Spillovers DD ¼ 0:05; aED ¼ aED ðG ¼ 15; y ¼ 0:3; aEE i 1 ¼ 0:1; ai 2 ¼ 0:2Þ.

Table 2.

Firm 2

Firm 1 EXP FDI  Nash equilibrium.

EXP

FDI

76.4, 76.4 77.8, 68.4

68.4, 77.8 69.1, 69.1

254

Table 3.

MARIA LUISA PETIT ET AL.

The Effect of Localized Spillovers on the Equilibrium Market Structure with Medium Spillovers DD ¼ 0:1; aED ¼ aED ðG ¼ 15; y ¼ 0:3; aEE i 1 ¼ 0:2; ai 2 ¼ 0:3Þ. Firm 2

Firm 1 EXP FDI

EXP

FDI

77, 77 78.7, 69.7

69.7, 78.7 70.1, 70.1

 Nash equilibrium.

Table 4.

The Effect of Localized Spillovers on the Equilibrium Market Structure with High Spillovers DD ¼ 0:2; aED ¼ aED ðG ¼ 15; y ¼ 0:3; aEE i 1 ¼ 0:3; ai 2 ¼ 0:4Þ. Firm 2

Firm 1 EXP FDI

EXP

FDI

78.2, 78.2 79.4, 70.8

70.8, 79.4 71, 71

 Nash equilibrium.

Table 5.

The Effect of the Degree of Spillover Localization (Parameters as in Table 2, but for aDD ¼ aED i 2 ¼ 0:4). Firm 2

Firm 1 EXP FDI

EXP

FDI

76.4, 76.4 81.3, 67.4

67, 81.3 72.7, 72.7

 Nash equilibrium.

Comparing Tables 2 and 5, we can observe that increasing the degree of spillover localization (i.e., increasing the value of the spillover parameter ED from aED 2 ¼ 0:2 up to a2 ¼ 0:4), improves the profits associated with the FDI choice (within a given market structure).

R&D and FDI with Asymmetries in Knowledge Transmission

255

4. CONCLUSIONS In this chapter, we examine the consequences of relaxing the assumption according to which firms generate and receive technological spillovers to the same extent. We argue that, on the contrary, firms may try to manage these information flows with the aim of maximizing incoming spillovers. First, firms may attempt at increasing the amount of external knowledge by means of a more efficient know-how management. Secondly, knowledge flows may increase due to proximity to rivals’ R&D labs if spillovers are localized. Taking into account these features we found that asymmetries in knowledge flows significantly affect firms’ innovative performance. In particular an increase in incoming spillovers intensity, due for instance to a better ability to manage knowledge flows, drives up the incentive to invest in own R&D. This result is in line with the findings of empirical research, indicating the existence of some complementarity between in-house R&D and external know-how. The relationship between the degree of asymmetry of knowledge flows, on one side, and the firms’ innovative performance, on the other side, becomes more complex if one considers an international oligopoly where both the optimal choice on international expansion and investment in R&D are endogenized. Taking into account the effects of geographical proximity on the transmission of know-how between firms may change the results obtained when these effects are ignored. In particular, if we compare two MNEs (the so-called DD case) with two exporters (the EE case), we find that whether exporters invest more in R&D than MNEs depends on the relative magnitude of two opposite forces. Since proximity implies a higher level of transmission of technological knowledge, the FDI–FDI choice gives rise to a free-riding effect, which is stronger than in the Export–Export case. If this effect prevails over the market expansion effect – due to the elimination of transport costs – firms may invest more in R&D when they both are exporters. However, a more clear-cut result emerges if one examines the so-called mixed duopoly (or ED case) where one firm chooses to serve the foreign market by exporting, and the rival becomes a multinational. We find that oneway FDI stimulates the multinational to raise its own R&D, due to both the elimination of transport costs and a greater ability to source. Furthermore, the model shows that the equilibrium R&D investment by the MNE raises after an increase in the degree of spillovers localization, whilst the investment of the exporting firm is hampered. This may be explained as a result of the increased advantage for the multinational firm deriving from its greater possibilities to source the rival’s knowledge through FDI.

256

MARIA LUISA PETIT ET AL.

As regards equilibrium market structures, we have shown that, increasing the degree of spillovers localization, a FDI–FDI equilibrium is more likely to occur. The possibility to absorb a higher proportion of the research produced by the competitor when investing in a foreign country appears to be a further incentive for firms to invest abroad. In fact, MNEs can make higher profits in relation to exporters (when compared with the case of non-localized spillovers) since, in a FDI–FDI equilibrium, MNEs have the possibility to reduce unit costs by free-riding on the research produced by the competitor.

NOTES 1. See also Piga and Poyago-Theotoky (2005), who present an Hotelling type model where firms competing in prices choose their R&D effort and the extent of location/product differentiation. 2. For instance, on the basis of survey data, Gupta and Govindarajan (2000) find a surprisingly low intensity of knowledge transfers within the MNE network (i.e., considering two-way knowledge transfers between parent and subsidiary and twoway knowledge flows among sister subsidiaries). We should expect the intensity of inter-firm spillovers – as we consider – to be lower than that of intra-firm spillovers – as examined by Gupta and Govindarajan. A similar conclusion can be reached by estimating an ‘‘innovation function’’ with inputs given by R&D activity carried out in a given region and R&D performed in other regions at different and increasing geographical distances, as in Bottazzi and Peri (2003). The econometric analysis considers the coefficients of the employment in R&D in other regions as a measure of the intensity of cross-regional spillovers. This comes out to be significant, decaying with distance and markedly lower than 0.5. 3. Petit and Sanna-Randaccio (2000) considers only the case of symmetric spillovers, and thus cannot address issues such as increased efficiency in know-how management or spillovers localization. Reference to localized spillovers can be found in Sanna-Randaccio (2002). 4. For a similar specification see D’Aspremont and Jacquemin (1988) and Wang and Blomstrom (1992), who assume y ¼ 1. The specification of cost-reducing innovation can easily be extended to the case of product innovation, where R&D investments shift the intercept of the demand curve upwards as in De Bondt, Sleuwaegen, and Veugelers (1988) and Veugelers and Vanden Houte (1990). 5. Market size is increasing in a and 1/b. 6. In particular, we assume that

I i þ ai I j  k ¼ I; II;

Ai y

q1;I 

i; j ¼ 1; 2 iaj; a  bq2;I ; b

q1;II 

qi;k 40

i ¼ 1; 2

a  bq2;II ; b

a4Ai 40

i ¼ 1; 2.

R&D and FDI with Asymmetries in Knowledge Transmission

257

7. This assumption is justified by d’Aspremont and Jacquemin (1988) by noting that the technological possibilities linking R&D inputs to innovative output do not display any economies of scale. 8. As explained in De Bondt and Henriques (1995), some firms – but not necessarily those with lower initial costs – may be more efficient in absorbing and managing the rivals’ innovative ideas. This occurs not because they themselves are more able to be innovative but rather because they are better in learning from the others. 9. Moreover, by evaluating the sign of the condition: @2 Pi =@I i @I j ¼ ð4y2 ð2aj  1Þðai  2ÞÞ=9b i; j ¼ 1; 2, we conclude that R&D investments are strategic substitutes (complements) if aio(W)0.5 and ajo(W)0.5. 10. Due to the symmetric nature of the model the DE case will not be examined. 11. In the case of exporting firms, given the presence of transport costs s, the condition for a firm to be active becomes a  A  sW0. It follows that 2(a  A)Ws and, therefore, the numerator of Eq. (19) is positive. 12. For a better understanding of our assumptions, we employ at this stage an index referring to each country and an index referring to each firm. 13. Notice that eliminating the country indexes is straightforward for exporting firms as they have only one plant located in their home country. 14. For instance Malerba (1992) indicates a rate of innovation with an average value of around 0.30 (thus y ¼ 0.30). 15. We recall that in a model with symmetric spillovers, a switch from the Export–Export to the FDI–FDI equilibrium is induced by a change in some parameters, that is by introducing a higher efficiency in research (i.e., a higher y), or a reduction in the value of the plant-specific fixed cost G.

REFERENCES Almeida, P. (1996). Knowledge sourcing by foreign MNEs: Patent citation analysis in the US semiconductor industry. Strategic Management Journal, 17(Winter Special Issue), 155–165. Bottazzi, L., & Peri, G. (2003). Innovation and spillovers in regions: Evidence from European patent data. European Economic Review, 47(4), 687–710. Branstetter, L. (2000). Is FDI a channel of knowledge spillovers? Evidence from Japan’s FDI in the United States. NBER Working Paper No. 8015, Cambridge, MA. Cassiman, B., & Veugelers, R. (2002). R&D cooperation and spillovers: Some empirical evidence from Belgium. American Economic Review, 92(4), 1169–1184. Cassiman, B., & Veugelers, R. (2004). Foreign subsidiaries as a channel of international technology diffusion: Some direct firm level evidence from Belgium. European Economic Review, 48(2), 455–476. Castellani, D., & Zanfei, A. (2006). Multinational firms, innovation and productivity. Northampton, MA: Edward Elgar. Cheng, L. (1984). International competition in R&D and technological leadership. Journal of International Economics, 17(1–2), 15–40. Cohen, W. M., & Levinthal, D. (1989). Innovation and learning: The two faces of R&D. Economic Journal, 99(3), 569–596.

258

MARIA LUISA PETIT ET AL.

De Bondt, R. (1997). Spillovers and innovative activities. International Journal of Industrial Organization, 15(1), 1–28. De Bondt, R., & Henriques, I. (1995). Strategic investment with asymmetric spillovers. Canadian Journal of Economics, 28(3), 656–674. De Bondt, R., Sleuwaegen, L., & Veugelers, R. (1988). Innovative strategic groups in multinational industries. European Economic Review, 32(4), 905–925. Duranton, G. (2000). Cumulative investment and spillovers in the formation of technological landscapes. Journal of Industrial Economics, 48(4), 127–139. D’Aspremont, C., & Jacquemin, A. (1988). Cooperative and noncooperative R&D in duopoly with spillovers. American Economic Review, 78(5), 1133–1138. Ethier, W. J., & Markusen, J. R. (1996). Multinational firms, technology diffusion and trade. Journal of International Economics, 41(1–2), 1–28. Fosfuri, A., & Motta, M. (1999). Multinationals without advantages. Scandinavian Journal of Economics, 101(4), 617–630. Frost, A. (1998). The geographic sources of innovation in the multinational enterprise: US subsidiaries and host country spillovers, 1980–1990. Ph.D. thesis, Sloan School of Management, MIT. Grunfeld, L. A. (2003). Meet me halfway but don’t rush: Absorptive capacity and strategic R&D investment revisited. International Journal of Industrial Organization, 21(8), 1091–1109. Gupta, A. K., & Govindarajan, V. (2000). Knowledge flows within multinational corporations. Strategic Management Journal, 21(3), 473–496. Jaffe, A. B., Trajtenberg, M., & Henderson, R. (1993). Geographic localization of knowledge spillovers as evidenced by patent citations. Quarterly Journal of Economics, 108(3), 577–598. Kamien, M. I., Muller, E., & Zang, I. (1992). Research joint ventures and R&D cartels. American Economic Review, 82(5), 1293–1307. Keller, W. (2002). Geographic localization of international technology diffusion. American Economic Review, 92(1), 120–142. Keller, W. (2004). International technology diffusion. Journal of Economic Literature, 43(2), 752–782. Malerba, F. (1992). Learning by doing and incremental technical change. Economic Journal, 15, 223–228. Neven, D., & Siotis, G. (1996). Technology sourcing and FDI in the EC: An empirical evaluation. International Journal of Industrial Organization, 14(5), 543–560. Petit, M. L., & Sanna-Randaccio, F. (2000). Endogenous R&D and foreign direct investment in international oligopolies. International Journal of Industrial Organization, 18(2), 339–367. Petit, M. L., Sanna-Randaccio, F., & Sestini, R. (2005). Localized spillovers and foreign direct investment: A dynamic analysis. DIS Technical Report no 7. University of Rome La Sapienza (pp. 1–23). Piga, C., & Poyago-Theotoky, J. (2005). Endogenous R&D, spillovers and locational choice. Regional Science and Urban Economics, 35(2), 127–139. Sanna-Randaccio, F. (2002). The impact of foreign direct investment on home and host countries with endogenous R&D. Review of International Economics, 10(2), 278–298. Singh, J. (2007). Asymmetry of knowledge spillovers between MNCs and host country firms. Journal of International Business Studies, 38(2), 764–786.

259

R&D and FDI with Asymmetries in Knowledge Transmission

Siotis, G. (1999). Foreign direct investment strategies and firm’s capabilities. Journal of Economics and Management Strategy, 8(2), 251–270. Veugelers, R. (1997). Internal R&D expenditure and external technology sourcing. Research Policy, 26(3), 303–315. Veugelers, R., & Vanden Houte, P. (1990). Domestic R&D in the presence of multinational enterprises. International Journal of Industrial Organization, 8(1), 1–15. Wang, J., & Blomstrom, M. (1992). Foreign investment and technology transfer. European Economic Review, 36, 137–155.

APPENDIX A. Equilibrium Outputs and Prices with Asymmetric Spillovers Equilibrium outputs resulting in the last stage of the game are given by ^DD q^DD i;I ¼ q i;II ¼

3gða  AÞ½3bg  4y2 ð2  ai Þð1  ai Þ C

i ¼ 1; 2

(A.1)

To guarantee the existence of a positive equilibrium solution for quantities, it must hold that ½3bg  4y2 ð2  ai Þð1  ai Þ40; i ¼ 1; 2: Notice that these very same conditions ensure also the positivity of equilibrium investments as given in (8). Taking into account equilibrium quantities it is then possible to calculate equilibrium prices ¼ p^DD p^DD I II ¼ a 

6bgða  AÞ½3bg  2y2 ðð2  a1 Þð1  a1 Þ þ ð2  a2 Þð1  a2 ÞÞ C (A.2)

B. Equilibrium Strategies for Sales and Equilibrium Prices (EE versus DD Duopoly) We begin with the case of a duopoly with two exporting firms, having that  sales of each firm in its own country: ~EE q~EE 1;I ¼ q 2;II ¼

3gða  AÞ 9bg  4ð2  aEE Þð1 þ aEE Þy2 þ

sð3bg  2ð2  aEE Þð1 þ aEE Þy2 Þ bð9bg  4ð2  aEE Þð1 þ aEE Þy2 Þ

ðA:3Þ

260

MARIA LUISA PETIT ET AL.

 sales of each firm abroad: ~EE q~EE 1;II ¼ q 2;I ¼

3gða  AÞ 9bg  4ð2  aEE Þð1 þ aEE Þy2 

2sð3bg  ð2  aEE Þð1 þ aEE Þy2 Þ bð9bg  4ð2  aEE Þð1 þ aEE Þy2 Þ

ðA:4Þ

We then substitute one of our assumptions on spillovers localization, i.e., aDD ¼ aDD ¼ aDD ; i; j ¼ 1; 2; into equilibrium outputs obtained at the last i j stage of the game in the DD case (see Eq. (A.1)). We get ~DD q~DD i;I ¼ q i;II ¼

3gða  AÞ 9bg  4ð2  aDD Þð1 þ aDD Þy2

i ¼ 1; 2

(A.5)

¼ aDD ¼ aDD 4aEE ¼ aEE ¼ aEE , the following proposiRecalling that aDD i j i j tion can be stated: Proposition 6. In an international duopoly with localized spillovers, sales by each firm both in its home country and abroad are higher if the firm is a MNE rather than an exporter. Proof of Proposition 6. Let us define YK ¼ ð3gða  AÞÞ=ð9bg  4ð2  aK Þ ð1 þ aK Þy2 Þ, with KA{DD, EE}. Therefore, taking into account sales in their own country by each firm in a MNE duopoly and in an exporting duoEE ~DD ~EE ~DD  YDD Þ  poly, respectively, we have that q~EE 1;I  q 1;I ¼ q 2;II  q 2;II ¼ ðY 2 EE EE EE EE 2 ðsð3bg  2y ð2  a Þð1 þ a ÞÞ=bð9bg  4ð2  a Þð1 þ a Þy ÞÞ: Focusing on the case of strategic substitutes (i.e., aKo0.5), we get that @YK =@a ¼ ð3gða  AÞ4y2 ð2a  1ÞÞ=ð9bg  4ð2  aÞð1 þ aÞy2 Þ2 40: Hence, ~DD ~EE ~DD q~EE 1;I oq 1;I e q 2;II oq 2;II . Likewise, considering sales abroad by each firm we DD DD ~DD ~EE can write q~1;I  q~EE  YEE Þ þ ð2sð3bg  2y2 ð2  1;II ¼ q 2;I  q 2;I ¼ ðY aEE Þð1 þ aEE ÞÞ=ðbð9bg  4ð2  aEE Þ ð1 þ aEE Þy2 ÞÞÞ40. & As a corollary of the above proposition, it is easily inferred that the level ~DD of aggregate production obtained for the DD case ðq~DD i;I þ q i;II ; i ¼ 1; 2Þ is larger than the corresponding aggregate level of sales obtained under the EE ~EE market structure ðq~EE i;I þ q i;II ; i ¼ 1; 2Þ.

R&D and FDI with Asymmetries in Knowledge Transmission

261

Equilibrium prices in the EE case, taking into account once again the hypotheses on spillovers localization, are given by 6bgða  AÞ 9bg  4ð2  aEE Þð1 þ aEE Þy2 3bgs þ 9bg  4ð2  aEE Þð1 þ aEE Þy2

~EE p~EE I ¼p II ¼ a 

ðA:6Þ

while in the DD case they are as follows: p~DD ¼ p~DD I II ¼ a 

6bgða  AÞ 9bg  4ð2  aDD Þð1 þ aDD Þy2

(A.7)

The outcome of a comparison of equilibrium prices is illustrated in the following proposition: Proposition 7. In an international duopoly with localized spillovers, equilibrium price is higher if both firms are exporters than when they both are MNEs. Proof of Proposition 7. Recalling that we defined YK ¼ ð3gða  AÞÞ= ð9bg  4ð2  aK Þð1 þ aK Þy2 Þ with K 2 fDD; EEg, and that, as already shown in the proof of Proposition 4, ðYDD  YEE Þ40, we have DD ~DD ~EE that p~DD  p~EE  YEE Þ  ð3bgsÞ=9bg  4ð2  I I ¼ p II  p II ¼ 2bðY 2 aEE Þ ð1 þ aEE Þy o0: &

This page intentionally left blank

FDI AND SPILLOVERS IN THE SWISS MANUFACTURING INDUSTRY: INTERACTION EFFECTS BETWEEN SPILLOVER MECHANISMS AND DOMESTIC ABSORPTIVE CAPACITIES Lamia Ben Hamida and Philippe Gugler ABSTRACT This chapter examines intra-industry spillover effects from inward foreign direct investment (FDI) in Swiss manufacturing firms. It suggests that (a) the assessment of spillovers calls upon a detailed analysis of these effects according to the mechanisms by which they occur (viz. the increase of competition, demonstration effects, and worker mobility), and (b) spillovers depend on the interaction between their mechanisms and the levels of domestic absorptive capacity. Results are affirmative in that high-technology firms benefit from FDI heightening competition, while mid-technology firms benefit from demonstration effects. And low-technology firms, which are not able to benefit from

New Perspectives in International Business Research Progress in International Business Research, Volume 3, 263–287 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1745-8862/doi:10.1016/S1745-8862(08)03012-4

263

264

LAMIA BEN HAMIDA AND PHILIPPE GUGLER

foreign affiliates via demonstration effects alone, manage to reap the benefit via the recruitment of MNCs labor. In addition, only firms which largely invest in absorbing foreign technology benefit from spillovers.

1. INTRODUCTION Foreign direct investment (FDI) is increasingly considered to be the main conduit of new technologies between countries – the creation, diffusion, and commercialization of technological innovations is one of the main characteristics of MNCs (Dunning & Gugler, 1994). It is argued that inward FDI is the principle source of positive spillovers for host economies (Dunning, 1992; Buckley, Clegg, & Wang, 2003). Many governments around the globe have liberalized their FDI regulations since the early 1980s and are now actively providing generous investment incentives to attract inward FDI (Dunning & Gugler, 2008; Dunning & Lundan, 2008; Oxelheim & Ghauri, 2003; UNCTAD, 2003). And the main motivation for these policies often stems from the expectation of FDI spillovers resulting in productivity enhancement of domestic firms. Generally, spillovers are said to take place when the entry and the presence of MNC affiliates lead to efficiency benefits in the host country’s local firms and the MNCs are not able to internalize the full value of these benefits (Blomstro¨m & Kokko, 1998). Spillovers are assumed to occur through four channels, viz. demonstration effects, competition effects, worker mobility, and backward–forward linkages. Although the effects via the last channel are also of a great importance and worthy to be explored, we focus in this contribution on studying the intra-industry spillovers. The number of empirical studies assessing the incidence of intra-industry spillovers to local firms is fast growing. Nonetheless, despite the policy relevance spillover effects of FDI on host economies are not well understood. So far, results have been mixed for both developed and developing countries and evidence on spillovers has not been conclusive yet.1 One of the reasons is that the share of foreign to total sectorial activity (e.g. foreign employment/ sales share) that has been by and large used by scholars to measure spillover benefits does not seem appropriate to capture much of competition (Kokko, 1996) and worker mobility (Ben Hamida, 2006a), it can only hold the whole information about spillovers through demonstration effects. Thereby, the assessment of spillover benefits needs to disentangle the effect of competition effects and worker mobility from that of demonstration effects by employing different control variables.

FDI and Spillovers in the Swiss Manufacturing Industry

265

Other reason for the apparently contradictory findings from the country studies is that domestic absorptive capacity may influence the incidence of spillovers (Wang & Blomstro¨m, 1992; Perez, 1998), in which only firms with high level of absorptive capacity are likely to benefit from FDI spillovers. The firm’s level of absorptive capacity depends upon its existing level of technological competence as well as its learning and investment efforts undertaken to be able to use productively foreign knowledge. This theoretical argument has been broadly taken into account by most empirical studies so as to be able to determine significant spillover effects. Nevertheless, these studies except Narula and Marin (2003) and Ben Hamida and Gugler (2006) disregard the importance of learning and investment efforts in determining the absorptive capacity of domestic firm and retain in most cases its existing level of technological capacity or its technological gap vis-a`-vis the foreign firm as proxies. Further possible explanation of these negative or insignificant spillover results is that the size and the extent of spillovers depend largely upon the interaction between the mechanisms by which they occur and the existing technological levels of domestic firms. Thereby, relatively high-technology firms are highly likely to benefit from spillovers through demonstration and/ or competition effects, whereas small technology firms which are not in position to compete with foreign firms, gain a lot from other forms of spillovers, such as worker mobility, since this channel provides some personnel assistance which can help domestic firms to better understand and implement the foreign technology (Mody, 1989). This contribution attempts to analyze empirically the intra-industry spillover effects from FDI using firm-level data from Swiss manufacturing industries. Switzerland is a particularly interesting example for this study given that it experiences increasing flows of inward FDI over time. It is regarded to have achieved competitive technological levels in many industries; MNCs tend to concentrate their activities in more dynamic and competitive industries. And Swiss government authorities (mostly at the cantonal level) are more and more active in attracting foreign MNCs. Our contribution differs from existing empirical literature with respect to three main points: first, it offers a more comprehensive picture of FDI intra-industry spillovers by distinguishing these effects according to their diverse channels. The share of total sales in the industry accounted for by foreign firms is used to capture the demonstration-imitation productivity effects while other control variables are used to assess the competition- and worker mobility-related spillovers. Second, it makes use of a thorough measure of domestic absorptive capacity in which the learning and

266

LAMIA BEN HAMIDA AND PHILIPPE GUGLER

investment efforts of domestic firms come with their existing technological capacities. Relatedly, it is argued that only domestic firms which largely invest in absorbing foreign technology benefit from spillovers. Third, it suggests that the size and the extent of spillovers depend largely upon the interaction between the mechanisms by which they occur and the existing technological capacities of domestic firms. The structure of the contribution is as follows. Following this introduction, Section 2 analyzes the framework underlying the empirical results, Section 3 presents the model, Section 4 discusses the Swiss data and some descriptive statistics, Section 5 presents the regression results, and Section 6 concludes the contribution.

2. INWARD FDI AND SPILLOVERS: THE FRAMEWORK As previously noted, MNCs are widely considered the main source for spillover benefits reflected in productivity of the domestic host country’s firms. In fact, MNCs are assumed to possess a countervailing advantage over the domestic firms in host countries (Hymer, 1960, 1968) since they use advanced technology in production, marketing, management, etc., which makes them more efficient than domestic firms (Dunning & Rugman, 1985). Such advanced technology may spill over to domestic firms allowing them to improve their performance. Positive spillovers represent the main elements justifying the effort made by government to attract foreign investors, although the expected potential benefits include among others, employment creation, capital formation, export promotion, etc. FDI intra-industry spillovers benefits are assumed to occur through three channels, viz. competition effects, worker mobility, and demonstration effects. Domestic firms may improve their productivity when the increase in competition that occurs as a result of foreign entry forces domestic firms to introduce new technology and/or work harder.2 When domestic workers already trained by or having worked in MNCs affiliates may decide to leave and join an existing or open up a new local firm, taking with them some or all of the firm-specific knowledge of the multinational. And when the foreign firms after entering the market demonstrate their advanced technologies and local entrepreneurs, after observing a product innovation or a novel form of organization adapted to local conditions, may recognize their feasibility and thus strive to imitate them (Meyer, 2003). In this way, knowledge is diffused as a result of the direct contact between foreign affiliates and local firms

FDI and Spillovers in the Swiss Manufacturing Industry

267

operating at different technological levels. Meyer adds that ‘‘the learning through observation affects not only technological innovation, but also new management techniques and new ways of inter-firm division of labor’’. A large literature has developed over the past two decades the concept of intra-industry spillovers. Too often, scholars in theoretical analysis offer a partial description of such spillovers, since each of them analyzes merely one kind of these effects. In Kopecky and Koizumi (1977), Findlay (1978), and Das (1987) spillovers are determined by the degree of foreign presence alone – contagion-type spillovers, measured for example in Findlay’s studies by the ratio of the capital stock of foreign-owned firms in the backward economy to the capital stock of the domestically owned firms. While, in Wang and Blomstro¨m (1992), Perez (1998), and Nakamura (2002), spillovers are rather endogenously generated by the technological competition between foreign affiliates and domestic firms – competitionrelated spillovers. And in Kaufmann (1997), Fosfuri, Motta, and Ronde (2001), and Glass and Saggi (2002) spillovers are the outcome of the movement of domestic labors who have been previously trained or worked at MNCs affiliates (Gugler & Brunner, 2007). Just as spillovers have not been analyzed at the theoretical level in a complete picture with respect to their diverse channels, so empirical studies are also focused on given partial analyses of these effects. In fact, spillover effects have been by and large measured by the share of foreign presence in the corresponding industry. This variable seems to be inappropriate to capture much of the competition- (Kokko, 1996) and worker mobility-related spillovers (Ben Hamida, 2006a). Even if the share of foreign to total sectorial activity seems to be an appropriate measure for spillover effects through demonstration, it cannot hold the whole information about competition and worker mobility effects. This is one of the reasons why there are evidence contrasts in the scant empirical evidence available. Yudayeva, Kozlov, Melentieva, and Ponomareva (2000), Castellani and Zanfei (2001), Haskel, Pereira, and Slaughter (2002), and Karpaty and Lundberg (2004) for example find positive evidence for the existence of spillover benefits from FDI for Russia, Italy, UK, and Sweden, respectively. While, using the same measure of spillovers, Aitken and Harrison (1999), Djankov and Hoekman (2000), and Castellani and Zanfei (2002) report negative and insignificant spillovers for Venezuela, Czech Republic, and Italy, respectively. If the share of foreign presence is not appropriate to assess the spillover effects from the increase of competition and the worker mobility, it seems clear that studies using this measure may yield misleading results. Thus, assessing the overall spillover effects needs to disentangle the effect of

268

LAMIA BEN HAMIDA AND PHILIPPE GUGLER

competition and worker mobility from that of demonstration by employing different control variables for each spillover mechanism. Doing so, the precise process of spilling-over will be correctly described in a more satisfactory model and then the impact of this process will be exactly identified. Given that, this contribution makes use of foreign sales share to assess the spillover effects from demonstration effects and employs other variables for competition effects and worker mobility-related spillovers.3 Other possible reason for those mixed results is that spillovers largely depend on the level of the absorptive capacity of domestic firms (Cantwell, 1989), in which only domestic firms possessing sufficient levels of absorptive capacity are likely to efficiently exploit spillovers, whereas insufficient absorptive capacity may hinder critical learning processes at the firm which in turn could not exploit the technological opportunities arising from foreign presence (Cohen & Levinthal, 1989). The diffusion of knowledge across borders may be limited because of the low absorptive capacity of potential recipients located abroad (Rugman & Verbeke, 2001). The concept of absorptive capacity encompasses the firm’s ability to recognize valuable new knowledge, integrate it into the firm and use it productively. Thereby, the firm’s level of absorptive capacity depends upon its existing level of technological competence as well as its learning and investment efforts undertaken to be able to use productively foreign knowledge. As suggested by Narula and Marin (2003), ‘‘absorption is not purely about imitation’’, in that technologies have a certain firm-specific aspect to them and then need to be decoded so as to be efficiently used by domestic firms raising their productivity. Thus, it is expected that only domestic firms which largely invest in absorbing foreign technologies benefit from FDI spillovers. Absorptive capacity has been broadly undertaken by most empirical studies so as to be able to determine significant spillover effects. Nevertheless, those studies disregard the importance of learning and investment efforts in determining the absorptive capacity of domestic firm and retain in most cases its existing level of technological capacity or its technological gap vis-a`-vis the foreign firm as proxies. Among others, Konings (1999), Girma, Greenaway, and Wakelin (1999), and Liu, Siler, Wang, and Wei (2000) used R&D intensity, technological gaps, and intangible assets per worker as proxies for domestic absorptive capacity. Using panel data on domestic firms of Bulgaria, Romania, and Poland, panel data on UK manufacturing firms, and panel data on UK manufacturing industries, Konings, Girma et al., and Liu et al., respectively, reported evidence of positive spillovers for R&D-intensive firms in Bulgaria and Poland, for all UK firms with low-technology gaps, and in UK

FDI and Spillovers in the Swiss Manufacturing Industry

269

industries with high technological capacities in terms of intangible assets. Early exceptions are Narula and Marin (2003) and Ben Hamida and Gugler (2006) who assert, respectively, that only Argentinean manufacturing firms and Swiss manufacturing firms that have invested more in absorptive capacities (in respect of investments in new equipment for product/process innovation and training activities) receive positive spillovers from FDI. Conversely, Narula and Marin conclude that the distinction of sectors according to different levels of technology gap does not provide any significant spillovers. Further reason for the negative or insignificant spillover results is that the size and the extent of spillovers depend largely upon the interaction between the mechanisms by which they occur and the existing technological levels of domestic firms. As stated by Mody (1989), relatively high-technology firms are highly likely to benefit from spillovers through demonstration and/or competition effects, whereas small technology firms which are not in position to compete with foreign firms, gain a lot from other forms of spillovers such as worker mobility, since this channel provides a (technical, managerial, etc.) assistance which can help domestic firms to better understand and implement the foreign technology. This shows that even low-technology firms may experience some spillover benefits from FDI and that only firms with very low technological competence to a point that they are not capable of reaping profits via any of the spillover channels enter into a process of cumulative decline and eventually leave the market. Recently, Ben Hamida (2006a) has analyzed in a theoretical paper FDI spillovers according to their diverse channels and found that the firm, which is not far behind the technological frontier of the industry, manages to exploit fully the technological opportunities using merely demonstration effects, whereas the firm of low technological development group is not able to benefit from foreign affiliates via demonstration effects alone, rather, it gains a lot from worker mobility. To test these theoretical findings, Ben Hamida (2006b) has used a qualitative method based on interviews with managers of foreign and domestic firms from Swiss manufacturing and service/construction. Doing so, the interviews analysis suggests that the theoretical findings remain pertinent for Switzerland. However, as qualitative analysis does not allow for the measurement of the size and the extent of spillover effects, we believe that further quantitative analysis would be promising. In this context, testing whether the increase in domestic productivity are function of the interaction effects between spillover channels and the technological capacities of domestic firms is the focal point of our empirical analysis discussed in this contribution.

270

LAMIA BEN HAMIDA AND PHILIPPE GUGLER

3. VARIABLES AND ECONOMETRIC SPECIFICATION We model the effects of FDI intra-industry spillovers within the context of a production function,4 in which the change in the natural log added value of the ith domestic firm is determined as follows: DLnY i;j ¼ a0 þ a1 DLnK i;j þ a2 DLnLi;j þ a3 FPj þ a4 HCi;j þ a5 FPj n HCi;j þ a6 DCompi;j þ a7 Sizei;j

(1)

þ a8 Industryj þ i;j where the subscripts i and j denote firm and industry, D represents changes in the variables between 2001 and 2004, and a0, a1, a2, a3, a4, a5, a6, a7, and a8 the parameters to be estimated. Table 1 describes the variables and their measurements. Y denotes added value at firm level, K its physical capital, L its employment, and HC the level of its human capital. The coefficients of those variables are expected to be positive and significant. Size, defined by the sales of firm i, is expected to increase productivity as larger sized firms may be more efficient (Dimelis & Louri, 2002). The inclusion of industry dummies,5 Industry, in Eq. (1) and the use of changes over the time control for the industry-specific productivity differences; they correct for the omission of unobservable variables that might undermine the relationship Table 1. Variables DLnY i;j DLnK i;j DLnLi;j FPj HCi;j DCompi;j Sizeij GAPij INVESTi,j

Variable Definitions. Definitions

The log change in added value at the firm level. The log change in physical capital, measured by gross capital income – firm level. The log change in total number of employees in a firm. The share of total sales in an industry j accounted for by foreign firms. The average labor cost of the firm (in 100,000 CHF) constructed as the ratio of the firm’s labor costs to the number of employees. The change in price markup at firm level measured by the difference between firm’s total sales and costs over total sales. The log total sales of the firm. The ratio of the average labor productivity of foreign-owned firms to domestic firm’s own labor productivity, calculated for 2001. The level of investment expenditures in new equipment and training activities for product/process innovation, within the period 2002–2004.

FDI and Spillovers in the Swiss Manufacturing Industry

271

between spillover variables and productivity growth of domestic firms (Aitken & Harrison, 1999; Narula & Marin, 2003). To assess the overall spillover effects of foreign firms on domestic competitors, we employ three different control variables with respect to the three possible intra-industry spillover mechanisms: first, the main effect6 of the share of foreign presence at four-digit industry level,7 FP, reflects spillovers from demonstration effects, resulting from the technology transfer that occurs from the direct contact between local agents and foreign affiliates operating at different levels of technology (Ben Hamida & Gugler, 2006). Second, the interaction term FP  HC between foreign presence and human capital is assumed to determine the effect of worker mobility related to the presence of foreign firms in the domestic market. In fact, this interaction assesses the combined effect of those variables on productivity of domestic firms; that is the influence of foreign firms would be co-determined by the level of human capital of the domestic firms. It is argued that human capital increases the ability of domestic firms to benefit from positive spillovers (Borensztein, De Gregorio, & Lee, 1998; Meyer & Sinani, 2001) – the sign of the interaction effect is then expected to be positive. Moreover, the technique of upgrading the level of the firm’s human capital depends on its existing technological level. On the one hand, relatively high-technology firms attempt to benefit from spillovers through demonstration and/or competition effects (Mody, 1989). Thereby, the ability of such firms to either absorb foreign technology or pursue independent lines of technological development, associated with the quality level of human capital, would be largely determined by the amount those firms spend in training their existing employees and/or the new ones so as to acquire the specific technique required either for the implementation of foreign knowledge or for the development of the existing one. On the other hand, small technology firms are not able to benefit from foreign affiliates via demonstration effects alone as they do not possess a sufficient level of human capital that allow them to exploit efficiently the foreign technological opportunities, rather they gain a lot from worker mobility, since this channel provides a (technical, managerial, etc.) assistance which can help them to better understand and implement the foreign technology. For that, to upgrade their level of human capital and then be able to use properly foreign best technology, those firms tend to invest in recruiting domestic employees already trained by or worked in foreign firms by giving them higher salary than foreign firms do8 – it is assumed that when leaving the MNCs those employees will take with them some or all of the firm-specific knowledge (Blomstro¨m & Kokko, 2002).

272

LAMIA BEN HAMIDA AND PHILIPPE GUGLER

Third, regarding competition-related spillovers, we use price markup or the so-called Lerner Index as a measure of competition – the difference between firm’s price (p) and its marginal cost (mc) over its total price. Lerner Index measures the degree to which firms can markup price above marginal cost; the larger the Lerner Index and the greater the power of the monopolist. The Lerner Index is also known as the Market Power index (Baye, 2006) as it describes the power a firm has within a market; e.g. a monopoly has the power to set high differences ( pmc) and so will have a high Lerner Index, while in a highly competitive market, each firm will have tight value of (pmc) and low Lerner Index. Unfortunately, the data set available do not allow for firm’s price and marginal cost informations. So, following Narula and Marin (2003) and Chung (2001) we use the difference between firm’s sales and its costs over its total sales as a measure of the firm’s price markup. When markup is high, a value near 1, competition is low. While, when markup is low, a value near 0, competition is high.9 As competition-related spillovers are associated with the increase in the level of competition that occurs as a result of foreign entry and presence, it seems more appropriate to take the change in markup to measure the change in the level of competition. A negative coefficient estimate attracted by the change in markup is consistent with the expectation that decreased markup (increased competition) is followed by productivity increase. To test the hypotheses, in which the size and the extent of spillover effects may vary according to the diverse levels of technological capacity of domestic firms and their absorptive capacity with respect to learning and investment efforts, we proceed to make various tests using Eq. (1). As a first step, we divide the full sample of domestic firms into three sub-samples characterized by the size of their existing technological capacities and estimate Eq. (1) separately for domestic firms with high, mid, and small technological capabilities. The existing technological capacities of domestic firms are measured by their technological gaps, GAP, compared to their foreign rivals. GAP is defined as the ratio of the average labour productivity of foreign-owned firms in the relevant four-digit industry to domestic firm’s own labor productivity, calculated for 2001. Hence, GAP is equal to one if the domestic firm operates at the same labour productivity as the average of its foreign rivals. Values that are smaller than or equal to one – the technological frontier of the industry – are interpreted as signs of small productivity gaps. Values which are higher than one but not far behind the technological frontier of the industry are interpreted as signs of mid productivity gaps, and those which are far behind the technological frontier

273

FDI and Spillovers in the Swiss Manufacturing Industry

characterize high productivity gaps. We expect to find stronger signs of competition-related spillovers in the sub-sample with small technological gap, whereas demonstration- and worker mobility-related benefits tend to take place in sub-samples of mid and high technological gap, respectively. As a second step, we divide the full sample into two sub-samples according to the investment level of domestic firms, INVEST, in the absorptive capacity. INVEST is measured by the level of investment expenditures in new equipment and training activities for product/process innovation, within the period 2002–2004. We expect that only domestic firms which largely invest in absorptive capacities benefit from FDI spillovers. We test for the equality of coefficients across sub-samples using Chow-tests. All results refer to OLS estimations of Eq. (1).

4. DATA AND DESCRIPTIVE STATISTICS Data for this contribution are derived from innovation activity surveys (2002 and 2005) of manufacturing firms, with at least five employees, conducted at the Swiss institute for business cycle research ‘‘KOF’’.10 Individual information covers the technological behavior and productivity performance of 1,201 firms – 185 majority-owned foreign affiliates – in 2001 and 1,134 firms – 182 majority-owned foreign affiliates – in 2004.11 Tables 2–6 present a summary of the samples and descriptive statistics of the relative position of foreign versus domestic firms. All these calculations are based on weighted data sets so as to give a representative picture of Swiss economy.12 As shown in Table 2, the share of foreign investment in manufacturing total employment accounted for 2001 is about 19 (21.6 in total sales). This share hides significant differences across sectors as shown in Table 3, in that 84 percent (94.7 in sales) of computer and office equipments is foreign Table 2. FDI Participation in Manufacturing in Switzerland: Annual Shares of Foreign Firms in Sales and Employment (Percent). Year

Total Employment

Total Sales

Number of Foreign Firms

Number of Domestic Firms

Total

2001 2004

19 17.6

21.6 19.6

185 182

1,016 952

1,201 1,134

Source: Author’s calculations of data derived from KOF innovation surveys (2002, 2005) of manufacturing and services/construction firms.

274

Table 3.

LAMIA BEN HAMIDA AND PHILIPPE GUGLER

FDI Participation in Manufacturing in Switzerland: Sectoral Shares of Foreign Firms (Percent).

Sector

Manufacturing Food Textiles Wood products Paper Printing and publishing Chemicals Pharmaceuticals Plastics Non-metal mineral products Metal production Metalworking Machinery Electrical machinery Computer and office equipments Communication equipments Medical instruments Watches Transport equipments Other manufacturing

Total Employment

Total Sales

2001

2004

2001

2004

13.3 13.8 9.5 32.1 2.5 25 13.2 20.6 16.9 6.9 12.9 28.9 26.4 84 15 20.1 5.1 33.2 15.9

4.3 14.9 5.4 25 8.3 22.2 13 23.7 11.3 11.9 10 22.9 49.7 11.6 40.1 27.1 2.2 24.8 4.4

15 16.5 25.3 38.3 8.8 21.8 7.1 29.1 15 10.9 17.7 32.4 31 94.7 13.5 21.8 9 43.9 21.7

2.9 13.6 6.5 29 12.7 25.6 23.5 32.4 13.4 13.8 13.9 21.3 59.3 11.4 54.1 35.4 0.7 23.2 8.1

Source: Author’s calculations of data derived from KOF innovation surveys (2002, 2005) of manufacturing and services/construction firms.

owned compared to only 2.5 percent (8.8 in sales) for printing and publishing. The foreign presence is also preeminent in among others paper, machinery, electrical machinery, and transport equipments. In spite of the slight decrease of the foreign employment and sales shares at the aggregate level from 2001 to 2004, there is a significant increase in foreign share across sectors. That is six of these sectors recognize a substantial increase in foreign employment share (eight in foreign sales share). Nonetheless, in other sectors foreign share falls by as much as 50 percent (such as food, watches, wood products, and computer and office equipments). Table 4 compares the relative performance of foreign and domestic firms across sectors in 2004, measured by total sales, total employment, and export as a percentage of total sales, using the ratio of foreign to domestic means. In general, the differences in the aggregate indicate that foreign manufacturing firms in Switzerland are larger than domestic firms, mainly in

275

FDI and Spillovers in the Swiss Manufacturing Industry

Table 4.

The Relative Position of Foreign Versus Domestic Firms: Sales, Labor, and Export (2004).

Ratio of the mean of the foreign variable to the mean of the corresponding domestic variable Sector

Sales

Labor

Export

Manufacturing

1.5

1.14

1.7

Food Beverage Textiles Wood products Paper Printing and publishing Chemicals Pharmaceuticals Plastics Non-metal mineral products Metal production Metalworking Machinery Electrical machinery Computer and office equipments Communication equipments Medical instruments Watches Transport equipments Other manufacturing

0.8 4 0.8 0.8 1.8 1.2 0.5 0.4 2.4 1.9 1.1 2.5 1.1 4 4.5 3.2 1.2 0.5 5.5 1.8

0.9 2 1 0.8 0.9 1.4 0.7 0.4 1 1 0.7 1.2 1.1 2.1 7.7 1.5 1.2 0.7 2.6 0.8

1 0.3 1.6 3.8 1 1.5 1.4 1.5 1.9 2 1.4 1.5 1.3 1.1 0.04 1.3 1.3 3.3 1.8 2.3

Two-sample t-test for equal means, which for simplicity does not take into account the sample design specificities. , , and  denote significance at the 10%, 5%, and 1% levels, respectively. Source: Author’s calculations of data derived from KOF innovation surveys (2002, 2005) of manufacturing and services/construction.

sales and export (about twice) – these differences are strongly significant with regard to export. This stems from the significant dominance of foreign firms, in terms of sales, mainly in sectors such as computer and office equipments, transport equipments, beverage and metalworking; in terms of export in wood products, watches, and other manufacturing; and in plastics regarding both sales and exports. Labor differences also favor foreign firms in some sectors such as communication equipments, and also computer and office equipments and transport equipments.13 In what concerns the relative technological position of foreign versus domestic firms, Table 5 reports the results of this comparison, in 2001 and

276

LAMIA BEN HAMIDA AND PHILIPPE GUGLER

Table 5. Affiliates’ Technological Behavior Relative to Domestic Firms: Labor Productivity, R&D Labor, and the Share of Innovative Products in Sales. Ratio of the mean of the foreign variable to the mean of the corresponding domestic variable Labor Productivity

Sector

2001

R&D Labor

Share of Innovative Products

2004

2001

2004

2001

2004

Manufacturing

1.2

1.2

1.3

1.4

1.1

1.1

Food Beverage Wood products Paper Printing and publishing Chemicals Pharmaceuticals Plastics Non-metal mineral products Metal production Metalworking Machinery Electrical machinery Computer and office equipments Communication equipments Medical instruments Watches Other manufacturing

1.8

1.1 2 0.9 1.2 1.2

1.9 0.2 0.6 0.6 8

1 3.3 0.7 0.4 8

0.9 0.2 0.4 0.6 1.3

0.7 0.1 0.05 1.5 1.7

2 1.3 1.2 1.6

0.5 0.9 0.3 1.1

0.6 2.3 0.9 0

1.1 1.3 1 0.6

1.3 1.4 1.5 0.6

2.2 2.5 1.3 2.2 1 1.3 1.2 1.2 1.3 1.1 1.2 1.1 1.4

1.2 1.3 1.1 1.4 1.2

0.2 0.8 1.3 1 1.4

1 0.3 1.4 1.6 0.2

1.5 0.8 1.1 0.9 1.5

0.2 0.7 0.9 0.9 0

1.1

1.5

0.6

1.9

1.1

1.1

1.1 1.6 1.3

1.5 0.26 2.1

1.2 1.7 0.2

0.6 0.5 3.4

1.1 2.4 0.7

0.9 1.5 1.8

Two-sample t-test for equal means, which for simplicity does not take into account the sample design specificities. , , and  denote significance at the 10%, 5%, and 1% levels, respectively. Source: Author’s calculations of data derived from KOF innovation surveys (2002, 2005) of manufacturing and services/construction.

2004, for the share of innovative products in sales, the share of R&D labor, and the labor productivity expressed as value added per employee. In 2001, the data for the aggregate suggest that on average foreign firms hire R&D employees and innovate more than domestic firms; the share of R&D labor of foreign firms is about once and half more than the share of R&D labor of

277

FDI and Spillovers in the Swiss Manufacturing Industry

Table 6.

Human Capital Development: Difference Between Foreign and Domestic Firms (2004).

Ratio of the mean of the foreign variable to the mean of the corresponding domestic variable Sector Manufacturing Food Beverage Textiles Wood products Paper Printing and publishing Chemicals Pharmaceuticals Plastics Non-metal mineral products Metal production Metalworking Machinery Electrical machinery Computer and office equipments Communication equipments Medical instruments Watches Transport equipments Other manufacturing

Professional Employees

Labor Quality

Labor Cost

1.1 1.2 1.3 0.9 1.1 1 1.1 1.1 1.1 1 1 1.1 1 1 1.1 1.4 1 1 1.4 1 1

1.7 3 2.1 0.7 0.8 1 2.6 3.2 2.2 0.9 0.7 0.9 0.9 1.4 1.2 7.5 0.7 1.2 1.1 1 0.7

1.2 1.1 1.1 1.1 1.2 1 1.1 1.1 1.1 1.2 1.1 1.1 1.1 1.2 1.3 1.3 1.3 1.1 0.4 1.5 1.3

Two-sample t-test for equal means, which for simplicity does not take into account the sample design specificities. , , and  denote significance at the 10%, 5%, and 1% levels, respectively. Source: Author’s calculations of data derived from KOF innovation surveys (2002, 2005) of manufacturing and services/construction.

domestic firms, which is significant at the 10% level. In spite of the slight change of these differences from 2001 to 2004, the result change considerably across sectors. That is some sectors recognize a substantial increase in favor of foreign firms mainly in beverage, pharmaceutical, and communication equipments in terms of R&D labor; and in paper and plastics in terms of innovative products. While in other sectors those differences markedly decrease highlighting domestic firms, such as in food, medical instruments, and watches in terms of R&D labor and innovative products.14 The difference in terms of productivity denotes the industrial technological gap between domestic and foreign firms, which is in favor of foreign firms and significant at the 1% level for 2001 at the aggregate level. This shows

278

LAMIA BEN HAMIDA AND PHILIPPE GUGLER

that on average the gap is relatively high – marks the relative productivity performance of foreign firms over their domestic rivals – and appears not to be associated with a catching-up process by domestic firms between 2001 and 2004. Nevertheless, when the sectors are considered individually these results change considerably and show that in some sectors the gap is small, while in others it is very high. In food and printing and publishing, for example, foreign firms perform better than domestic ones, while in chemicals domestic firms are found to be at the same technological level as foreign affiliates. Moreover, beverage experiences a large gap in 2001 associated with a process of falling behind in 2004, whereas food, wood products, and watches succeed in catching-up with and even in forging ahead of foreign firms. This catchingup process may result from the investment effort of domestic firms in learning activity as, for example, in wood product, machinery, and watches sectors the decrease in the technological gap appear jointly related to the increase of the share of innovative products. Whether at least some of the increases in productivity are due to spillover benefits arising from the learning process of foreign technologies is the focal point of our empirical analysis discussed in the next section.15 Finally, Table 6 analyzes the relative contribution of foreign firms to domestic human capital development versus domestic counterparts, in 2004. Variables used are the share of professionals – engineers, mangers, and all other professionals in production and R&D activities – in total employees, the labor quality index expressed in terms of the ratio of professional to non-professionals, and the share of labor cost in sales (including salary, training expenditures, etc.). The data for the aggregate suggest that foreign firms hire more professionals which for the most part consist on R&D employees (Table 5), the quality of their labor force is significantly higher, and they invest more in labor costs. The high level of labor cost perceived in foreign affiliates relative to similar domestic firms may result from the large amount they spend in training.16 This way, MNC affiliates may be particularly valuable sources of new technology and hence more opportunities for spillover benefits are expected. As suggested by Blomstro¨m and Kokko (2002), the labor market is one of the main ways in which new technological knowledge is expected to disseminate to the domestic economy, workers already trained by or worked in foreign affiliates may be potentially available to work in domestic firms or start their own firms in the same industry. In this respect, we find that many relatively small technology firms in 2001, which spend as much as or even more than foreign firms in labor cost, experienced in 2004 an increasing level of their technological development.17 This could be explained by the fact that

FDI and Spillovers in the Swiss Manufacturing Industry

279

those firms succeeded in attracting skilled domestic employees worked in foreign firms, qualified as appropriate to their productivity enhancement. Across sectors, the results in Table 6 show that foreign firms hire more professionals and possess a more skilled labour force in sectors such as chemicals; and invest more in labor costs in plastics, transport equipments, and electrical machinery.

5. RESULTS Regression estimates, column 1 in Table 7 shows the results of the spillover tests of the full sample of 370 Swiss manufacturing firms. The value added of the firms in Switzerland for the full samples increases with changes in the employment and the human capital of domestic firms. However, as expected, the estimated coefficient of the variable FP is negative and insignificant showing that foreign presence does not have any effect on productivity growth of domestic firms; so on average there is no evidence of technological spillovers from demonstration effects. The interaction term between FP and HC is also insignificant, indicating that the full sample data have not demonstrated the change in response with FP depends on the level of human capital. Similarly, the increase in competition seems to impede the productivity growth of domestic firms as the DComp estimate is positive and highly significant. And the physical capital and Size do not affect significantly the productivity change of domestic firms. In columns 2, 3, and 4 of Tables 7, we have proceeded to divide the sample of manufacturing into three sub-samples characterized by the values for the variable GAP. The results suggest that the estimated coefficients of FP and FP  HC are only positive and significant in the sub-samples of firms with mid and large technological gaps – when GAP is greater than one. Both kinds of firms manage to exploit fully the technological opportunities arising from their direct contact with foreign firms – demonstration-related spillovers. The size of such benefits is 0.009 for mid-technology firms, while 0.005 for low-technology ones, implying that an increase in the share of foreign investment from 0 to 10 percent leads to as much as 0.05 percentpoint increase in domestic productivity of low-level group and about twice larger for mid-level group.18 Mid- and low-technology firms also gain benefits from FDI by investing in human capital; the amount those firms spend in training their existing employees and/or the new ones appears to be of great importance for the successful implementation of foreign knowledge. The positive and significant interaction effects of FP with HC indicate that the effect of foreign firms is broadly co-determined by the level of human

280

LAMIA BEN HAMIDA AND PHILIPPE GUGLER

Table 7. Estimation Results for Manufacturing: Spillovers from FDI and the Level of Absorptive Capacity of Domestic Firms.

Variables

DLnK DLnL HC FPj FPnj HC DComp Size 2 R~ F–Chow N

1

2

3

4

5

6

Full

Small

Mid

Large

High

Small

0.0004 (0.004) 0.77 (0.07) 0.42 (0.06) 0.0002 (0.0009) 0.006 (0.004) 1.54 (0.14) 0.001 (0.008) 0.67

GAP 0.45 (0.04) 0.38 (0.1) 0.23 (0.07) 0.0005 (0.001) 0.00007 0.003 0.36 0.19 0.03 (0.01) 0.88

GAP 0.005 (0.006) 0.71 (0.09) 0.57 (0.1) 0.009 (0.002) 0.01 (0.006) 1.76 (0.15) 0.01 (0.01) 0.69

GAP 0.006 (0.005) 0.79 (0.05) 0.66 (0.07) 0.005 (0.001) 0.011 (0.003) 1.43 (0.1) 0.007 (0.01) 0.77

INVEST 0.002 (0.005) 0.95 (0.07) 0.41 (0.08) 0.004 (0.001) 0.005 0.003) 1.67 (0.1) 0.004 (0.01) 0.79

INVEST 0.006 (0.01) 0.67 (0.1) 0.37 (0.2) –0.002 (0.006) 0.004 (0.02) 1.45 (0.2) 0.02 (0.03) 0.6

370

71

15.6 106

193

8.16 179

61

Notes: All estimations include industry dummies. All standard errors, in parentheses, are corrected for heteroskedasticiy. Variables (HC and FP) used for interactions are centered by subtracting the full sample means, so that (1) multicollinearity between the variables and their product is reduced, (2) better estimates of HC and FP are ensured, and (3) more meaningful interpretations of those estimates are granted (Aiken & West, 1991). , , and  denote significance at the 10%, 5%, and 1% levels, respectively.

capital of the domestic firms – this finding confirms the strong association between FDI effects and the level of domestic human capital. Moreover, as we have mentioned in previous section, domestic firms and especially lowtechnology ones tend to upgrade their level of human capital by recruiting domestic employees already trained by or worked in foreign firms. Doing so, low-technology firms may get hold of some personnel assistance, essential to be able to decode and use effectively foreign best technology. In this respect, the positive and highly significant interaction effect of FP with HC – column 4 of Table 7 – could be a sign of worker mobility-related spillovers. This result seems consistent with Ben Hamida’s (2006a) theoretical analysis. The estimated coefficient of DComp is negative and significant only for the sub-sample of domestic firms with small technology gap, suggesting that

FDI and Spillovers in the Swiss Manufacturing Industry

281

heightened competition (decreased markup) is followed by productivity increase – mid- and low-technology firms do not benefit from the competition-related spillovers. The estimated coefficients of FP and FP  HC are insignificant for the high-technology manufacturing firms. This is not surprising given that high-technology firms do not need to learn from foreign technologies to increase their productivity since they perform as much as or even better than foreign rivals in the industry. Instead, those firms gain benefit from FDI via competition effects; the competitive pressure generated by the presence of foreign firms induces it to use more efficiently its existing technology by learning within its existing line of technological development. The estimated coefficients of HC are positive and significant for all subsamples with larger effects in mid- and low-technology firms. This can be explained by the fact that those kinds of firms substantially invest in upgrading their human capital to fully exploit the technological opportunities arising from foreign presence; in turn this investment effort has a great impact on their productivity growth. Columns 5 and 6 of Table 7 report the results of spillovers according to the level of the absorptive capacity in terms of learning and investment efforts for manufacturing firms. Estimated coefficients of both FP and FP  HC are positive and significant only for high-INVEST sub-samples, indicating that only domestic firms which highly invest in the absorption of foreign knowledge have more efficiently internalize FDI spillovers from technology transfer.19 The estimated coefficient of DComp for high-INVEST sub-sample is significantly positive, suggesting that there is no evidence for competition-related spillovers. Moreover, domestic firms which little invest in the absorptive capacity are not capable of reaping profits via any of the spillover channels. This findings confirm the importance of the investment and training efforts of domestic firms in productively absorbing foreign knowledge occurring from demonstration effects and worker mobility since both channels require further investment once introducing the foreign best technology in their existing technological process. The Chow-tests soundly support the divisions of manufacturing sample with respect to GAP and INVEST

6. CONCLUSIONS This contribution suggests that firstly, the assessment of the effects of spillovers from FDI on the productivity development of domestic firms calls upon a detailed analysis of those effects according to their ways of

282

LAMIA BEN HAMIDA AND PHILIPPE GUGLER

occurrence; and secondly, spillover benefits are determined by the interaction between the channels by which they occur and the technological characteristics of the recipient host firms. In this respect, we focus on testing the effects of the diverse intra-industry spillover channels according to the level of the absorptive capability of domestic firms. On the basis of samples of Swiss manufacturing, we show that it is important to take account of the level of technological capacity of the domestic firms as well as their investment effort in the absorptive capability when evaluating productivity spillovers generated from FDI. That is, taking all the firms together the results do not report on average significant evidence for spillovers, neither from the increase of competition nor from the technology transfer. However, looking separately at three sub-samples of firms characterized by the size of the technological gap between domestic and foreign firms, yields differences in results. Domestic firms with high technological capacity appear to benefit from spillovers which are basically from the FDI heightening competition. Mid-technology firms benefit a lot from demonstration effects, while low-technology firms which are not able to benefit from foreign affiliates via demonstration effects alone, manage to reap the spillover benefits via the recruitment of MNCs’ labor that can help them to successfully imitate foreign knowledge. Furthermore, when taking into account the investment level of domestic firms in the absorptive capacity, we find evidence for positive spillovers only in the sub-sample of firms with relatively high-investment level. Those benefits result from the FDI technology transfer. Spillovers, however, affect negatively the productivity of domestic firms which do not actively engage in investment and learning to be able to absorb foreign knowledge. On the policy front, suggestions with respect to attracting FDI following such findings must take into account that benefits from FDI in terms of spillovers require sufficient level of human capital, especially for mid- and low-gap firms, to be able to use efficiently foreign knowledge. In this respect, actions to motivate subsidization of foreign investment as well as to support learning and investment in domestic firms seem to be necessary ingredients in a policy package to maximize the technological spillovers from FDI. Future research aiming at analyzing foreign characteristics (such as the degree of foreign ownership, the nationality of foreign investors, the complexity level of MNC technology, the motivations for FDI, etc.) as determinants of spillover effects could be also of a great importance to policy-makers in leveraging the potential benefits of inward FDI spillovers.

FDI and Spillovers in the Swiss Manufacturing Industry

283

Also, exploring other kind of spillovers, such as inter-industry and market access spillovers could be promising.

NOTES 1. A meta-analysis of spillover studies is presented in Meyer and Sinani (2005). 2. Note that MNCs tend to concentrate their activities in competitive industries (Rugman & Verbeke, 2003). 3. Spillover variables will be exactly defined in the next section. 4. The derivation of this model is explained in the annex. 5. There are 32 industry dummies accounted for manufacturing. 6. It is also called the average effect (Aiken & West, 1991) since it denotes the effects of the FP on domestic productivity at the mean of HC as those variables used for interaction are centered (more details are given in footnotes to Table 7). 7. We use the maximum available disaggregation industry level to be able to effectively assess the intra-industry spillover benefits. 8. Foreign affiliates are unlikely to be mute spectators as their employees move to domestic competitors taking with them their secrets. 9. Note that in some cases a higher markup may be due to industry specificities as for example in luxury industry (Narula & Marin, 2003). 10. Questionnaires can be downloaded from www.kof.ethz.ch (Industrieo¨konomik). 11. Unfortunately, data of 2004 are the more recent ones. 12. The weights are used to correct for the selection bias resulting from ‘‘unit’’ non-response and for the deviations of the sample structure from that of the underlying population. 13. As indicated by the ratio of the mean of the foreign variable to the mean of the corresponding domestic variable, there exist sectors wherein domestic firms dominate foreign affiliates but these results are not significant. 14. Note that in some cases a higher markup may be due to industry specificities as for example in luxury industry (Narula & Marin, 2003). 15. The regression analysis make use of a sample of only 370 manufacturing firms because of missing data for some variables when matching the two data sets of 2002 and 2005 surveys. 16. Chen (1983) and Gershenberg (1987) found evidence that foreign firms spend more in training than domestic firms in Kenya and Hong Kong, respectively. 17. Domestic firms of industries such as watches succeeded in catching-up with foreign rivals. 18. Comparing with Ben Hamida and Gugler (2006)’s regression results of demonstration-related spillovers for manufacturing, the effect of FP in 2001 on the productivity change of mid-technological firms between 2001 and 2004 is smaller than that of 1998 on the change between 1998 and 2001. 19. This result seems consistent with Narula and Marin (2003)’s analysis.

284

LAMIA BEN HAMIDA AND PHILIPPE GUGLER

ACKNOWLEDGEMENT We thank Laurent Donze´ for his helpful comments and suggestions as well as Arvanitis Spyros and Nora Sydow for their data and econometric assistance.

REFERENCES Aiken, L., & West, S. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park, London: Sage Publications. Aitken, B. J., & Harrison, A. E. (1999). Do domestic firms benefit from direct foreign investment? Evidence from Venezuela. American Economic Review, 89, 605–618. Baye, M. R. (2006). Managerial economics and business strategy (5th ed). New York: McGraw-Hill. Ben Hamida, L. (2006a). Multinational firms, spillovers, and productivity growth: An evolutionary model. In: Proceeding of EcoMod international conference, regional and urban modeling, June 1–3, Brussels. Ben Hamida, L. (2006b). Intra-Industry spillovers from inward FDI: An evolutionary model. In: Proceedings of EIBA conference, regional and national divers of business location and competitiveness, December 7–9, Fribourg. Ben Hamida, L., & Gugler, P. (2006). Are there demonstration-related spillovers from FDI? Evidence from Switzerland. In: Proceedings of EIBA conference, regional and national divers of business location and competitiveness, December 7–9, Fribourg. Blomstro¨m, M., & Kokko, A. (1998). Multinational corporations and spillovers. Journal of Economic Surveys, 12(3), 247–277. Blomstro¨m, M., & Kokko, A. (2002). FDI and human capital: A research agenda. Working Paper, No. 195, OECD, Paris. Borensztein, E., De Gregorio, J., & Lee, J.-W. (1998). How does foreign direct investment affect economic growth? Journal of International Economics, 45, 115–135. Buckley, P. J., Clegg, J., & Wang, C. (2003). Is the relationship between inward FDI and spillover effects linear? An Empirical examination of the case of China. Presented at the 29th annual conference of European academy of international business, Copenhagen. Cantwell, J. (1989). Technological innovation and multinational corporations. Oxford: Basil Blackwell. Castellani, D., & Zanfei, A. (2001). Technology gaps, absorptive capacity and the impact of inward investment on productivity of European firms. Mimeo, ISE-Universita di Urbino. Castellani, D., & Zanfei, A. (2002). Multinational companies and productivity spillovers: Is there a specification error. Mimeo, ISE-Universita di Urbino. Chen, E. K. Y. (1983). Multinational corporations, technology and employment. London: Macmillan. Chung, W. (2001). Identifying technology transfer in foreign direct investment: Influence of industry conditions and investing firm motives. Journal of International Business Studies, 32, 211–229. Cohen, W., & Levinthal, D. (1989). Innovation and learning: The two faces of R&D. Economic Journal, 99, 569–596.

FDI and Spillovers in the Swiss Manufacturing Industry

285

Das, S. (1987). Externalities, and technology transfer through multinational corporations. Journal of International Economics, 22, 171–182. Dimelis, S., & Louri, H. (2002). Foreign investment and efficiency benefits: A conditional quantile analysis. Oxford Economic Papers, 54, 449–469. Djankov, S., & Hoekman, B. (2000). Foreign investment and productivity growth in Czech enterprises. World Bank Economic Review, 14, 49–64. Dunning, J. H. (1992). Multinational enterprises and the global economy. Wokingham, England: Addison-Wesley Publishing Company. Dunning, J. H., & Gugler, P. (1994). Technology based cross-border alliances. In: J. J. David (Ed.), Technology and business enterprise. Cheltenham: Edward Elgar, UK. Dunning, J. H., & Gugler, P. (2008). Foreign direct investment, location and competitiveness. Oxford: Elsevier. Dunning, J. H., & Lundan, S. (2008). Multinational enterprises and the global economy (2nd ed.). Cheltenham, UK, Northampton, MA, USA: Edward Elgar. Dunning, J. H., & Rugman, A. M. (1985). The influence of Hymer’s dissertation on the theory of foreign direct investment. The American Economic Review, 75, 228–232. Findlay, R. (1978). Relative backwardness, direct foreign investment, and the transfer of technology: A simple dynamic model. Quarterly Journal of Economics, 92, 1–16. Fosfuri, A., Motta, M., & Ronde, T. (2001). Foreign direct investment and spillovers through workers’ mobility. Journal of International Economics, 53, 205–222. Gershenberg, I. (1987). The training and spread of managerial know-how: A comparative analysis of multinational and other firms in Kenya. World Development, 15, 931–939. Girma, S., Greenaway, D., & Wakelin, K. (1999). Wages, productivity and foreign ownership in UK manufacturing. Working Paper, No. 99/14. University of Nottingham. Glass, A., & Saggi, K. (2002). Multinational firms and technology transfer. Scandinavian Journal of Economics, 104, 495–514. Gugler, P., & Brunner, S. (2007). FDI effect on national competitiveness: A cluster approach. International Advances in Economic Research, 13, 268–284. Haskel, J. E., Pereira, S. C., & Slaughter, M. J. (2002). Does inward foreign investment boost the productivity of domestic firms? Working Paper, No. 8724, National Bureau of Economic Research, Cambridge, MA. Hymer, S. (1968). La Grande firme multinationale. Revue Economique, 14, 949–973, Translated in English in Casson, M. C. (ed.), Multinational corporations, Edward Elgar, Cheltenham. Hymer, S. H. (1960). The international operations of national firms: A study of direct foreign investment. Ph.D. dissertation, Published by MIT Press, Cambridge (also published under same title in 1976), Karpaty, P., & Lundberg, L. (2004). Foreign direct investment and productivity spillovers in Swedish manufacturing. Working Paper, No. 2, Orebro University, Sweden. Kaufmann, L. (1997). A model of spillovers through labor recruitment. International Economic Journal, 11, 13–33. Kokko, A. (1996). Productivity spillovers from competition between local firms and foreign affiliates. Journal of International Development, 8, 517–530. Konings, J. (1999). The effects of foreign direct investment on domestic firms: Evidence from firm level panel data in emerging economics. LICOS Discussion Paper DP 8699. Kopecky, K. J., & Koizumi, T. (1977). Economic growth, capital movements and the international transfer of technical knowledge. Journal of International Economics, 7, 45–65.

286

LAMIA BEN HAMIDA AND PHILIPPE GUGLER

Liu, X., Siler, P., Wang, C., & Wei, Y. (2000). Productivity spillovers from foreign direct investment: Evidence from UK industry level panel data. Journal of International Business Studies, 31, 407–425. Meyer, K. (2003). FDI spillovers in emerging markets: A literature review and new perspectives. Working Paper, No. 15, Copenhagen Business School. Meyer, K., & Sinani, E. (2001). Identifying spillovers of technology transfer from FDI: The case of Estonia. Working Paper, Center for East European Studies, Copenhagen Business School. Meyer, K., & Sinani, E. (2005). Spillovers from foreign direct investment: A meta-analysis. Working Paper, Department of Economics, University of Reading. Mody, A. (1989). Strategies for developing information industries. In: C. Cooper & R. Kaplinsky (Eds), Technology and development in the third industrial revolution. London: Frank Cass. Nakamura, T. (2002). Foreign investment, technology transfer, and the technology gap: A note. Review of Development Economics, 6, 39–47. Narula, R., & Marin, A. (2003). FDI spillovers, absorptive capacities and human capital development: Evidence from Argentina. Working Paper, No. 2003-016, Maastricht Economic Research Institute on Innovation and Technology, The Netherlands. Oxelheim, L., & Ghauri, P. (2003). European union and the race for foreign direct investment in Europe. Oxford: Elsevier. Perez, T. (1998). Foreign investment and spillovers. the Netherlands: Harwood Academic Publishers. Rugman, A. M., & Verbeke, A. (2001). Subsidiary specific advantages in multinational enterprises. Strategic Management Journal, 22, 237–250. Rugman, A. M., & Verbeke, A. (2003). Multinational enterprises and clusters: An organizing framework. Management International Review, 43(Special Issue 3), 151–169. UNCTAD. (2003). World investment report, FDI policies for development: National and international perspectives. New York and Geneva: United Nation. Wang, J.-Y., & Blomstro¨m, M. (1992). Foreign investment and technology transfer: A simple model. European Economic Review, 36, 137–155. Yudayeva, K., Kozlov, K., Melentieva, N, & Ponomareva, N. (2000). Does foreign ownership matter? Russian experience. Working Paper, No. w0005, Center for Economic and Financial Research, Russia.

ANNEX: THE MODEL Eq. (1) is derived from a Cobb–Douglas production function with added value Y a function of two inputs, capital and labor a2 Y i;t ¼ Ai La1 i;t K i;t

(A.1)

The level of productivity is given by Ai;t , which is assumed to vary across firms within each sector j and across time t. After taking logarithms of variables to get into a linear form Eq. (A.1) and adding a stochastic disturbance term ui,t to account for variations in the

FDI and Spillovers in the Swiss Manufacturing Industry

287

productive capabilities of the ith firm, we can rewrite Eq. (A.1) for t–3 ¼ 2001 and t ¼ 2004 as LnY i;t ¼ ait þ a1 LnLi;t þ a2 LnK i;t þ ui;t ; ðai;t ¼ LnAi;t Þ

(A.2)

LnY i;t3 ¼ ai;t3 þ a1 LnLi;t3 þ a2 LnK i;t3 þ ui;t3 ; ðai;t3 ¼ LnAi;t3 Þ (A.3) Then, taking the difference (A.2–A.3) yields the change in value-added for domestic firms between 2004 and 2001. D denotes the variation between 2004 and 2001. DLnY i ¼ Dai þ a1 DLnLi þ a2 DLnK i þ i

(A.4)

We test the hypothesis that productivity growth is affected by the share of foreign presence at the industry level, its interaction with human capital of the ith firm, and the increase in the level of industry competition, by modeling the change in a as Dai ¼ a3 FPj;t3 þ a4 HCi;j;t þ a5 FPj;t3 n HCi;j;t þ DCompj þ a7 Sizei;j;t þ a3 Industryi;j

ðA:5Þ

where the change in a is also assumed to vary across sectors, the human capital of the domestic firm, and its size. Finally, combining Eqs. (A.4) and (A.5) yields Eq. (1).

This page intentionally left blank

MARKET AND TECHNOLOGY ACCESS THROUGH FIRM ACQUISITIONS: BEYOND ONE SIZE FITS ALL Christoph Grimpe and Katrin Hussinger ABSTRACT Purpose – Firm acquisitions have been shown to serve as a way to gain access to international markets, technological assets, products or other valuable resources of the target firm. Given this heterogeneity of takeover motivations and the skewness of the distribution of the deal value we show whether and how the importance of different takeover motivations changes along the deal value distribution. Methodology/approach – On the basis of a comprehensive dataset of 652 European mergers and acquisitions in the period from 1997 to 2003, we use quantile regressions to decompose the deal value at different points of its distribution. Findings – Our results indicate that the importance of technological assets is higher for smaller target firms while the importance of nontechnological assets seems to be higher for larger targets. The findings support the view on small acquisition targets to complement the acquirer’s

New Perspectives in International Business Research Progress in International Business Research, Volume 3, 289–314 Copyright r 2008 by Emerald Group Publishing Limited All rights of reproduction in any form reserved ISSN: 1745-8862/doi:10.1016/S1745-8862(08)03013-6

289

290

CHRISTOPH GRIMPE AND KATRIN HUSSINGER

technology portfolio while larger acquisition targets tend to be used to gain access to international markets. Research limitations/implications (if applicable) – Our findings suggest that the average firm as a reference for study might not be appropriate to address as the size of the target firm influences the value attribution to the target’s assets. Practical implications (if applicable) – Managers in the acquiring firm should be aware that they might overpay for the technological assets of a small firm. However, the acquisition of larger targets requires a well-developed integration strategy. Originality/value of paper – For the first time, the broad merger motive of technology acquisition has been further qualified according to the size of the target which exhibits a considerable impact.

INTRODUCTION The year 2007 marked a new record high for worldwide merger and acquisition (M&A) activity with more than 3.3 trillion Euros in total deal value spread over 59,000 transactions.1 While mega-mergers are closely observed by the public, the majority of acquisitions that target at small- and medium-sized enterprises (SMEs) receives almost no attention. This is surprising as acquisitions of small firms have been shown to be highly important in many industries, for example, in the beer industry (Sutton, 1991), but also in high-technology sectors like biotechnology (Graff, Rausser, & Small, 2003). Whereas small firms in the beer industry are attractive because of their retail outlets, SMEs in high-tech sectors are attractive acquisition targets because of their technological assets and intellectual property rights. In fact, the distribution of the deal value appears to be highly skewed: the vast majority of deals is rather small. Fig. 1 shows a histogram for the deal value for more than 8,300 M&A transactions in which either acquiring or target firms from the EU-27 were involved in 2007. Merger objectives have been shown to range from gaining access to international markets (e.g., Dunning, 1988), accessing technological assets and intellectual property rights (e.g., Ahuja & Katila, 2001; Graebner, 2004), or to buying certain product lines and other valuable resources of the target firm (Griliches, 1981; Pakes, 1985). The heterogeneity of merger objectives along with the skewness of the deal value distribution suggests,

291

0

.05

Density

.1

.15

Market and Technology Access through Firm Acquisitions

0

200

400

600

800

1000

Deal value (million Euro)

Fig. 1.

Skewness of the Deal Value Distribution.

however, that there is no common story behind M&A objectives. Yet, previous literature has failed to expose different takeover motives behind smaller and larger deals. Standard regression techniques usually focus on the ‘‘average firm’’, implicitly assuming that there is a common story to tell about all transactions. When it comes to investigating merger objectives it is questionable, however, whether this can provide the full picture. If an acquisition was targeted at particular intellectual property rights (O’Donoghue, Scotchmer, & Thisse, 1998; Lerner, Tirole, & Strojwas, 2003) or an innovative product line (Black, 2000) we would expect this to be achieved rather through a smaller acquisition target as takeovers can be more attractive in such cases than technology licensing agreements because the costs of governance are likely to be smaller than the costs of licensing (hold-up problems, value allocation problems, etc.) (Graff et al., 2003). Moreover, smaller firms tend to exhibit a higher level of creativity and innovativeness which, for example large pharmaceutical firms wish to exploit when acquiring small and innovative biotechnology firms (Teece, 1988). After all, they also seem to be easier to integrate (Grimpe, 2007). Contrary to this, acquisition targets can be big players in the market

292

CHRISTOPH GRIMPE AND KATRIN HUSSINGER

that provide the acquiring firm with the entry to a foreign market, with a large market share, product portfolio and a well developed distribution and customer network (Caves, 1989; Scherer & Ross, 1990; Ro¨ller, Stenneck, & Verboven, 2001). Moreover, the acquisition of a large player in the market immediately reduces competition. Given the intention to quickly enter a foreign market, acquisitions associated with a high deal value are hence likely to be cross-border rather than domestic (Dunning, 1988). Given the heterogeneity of M&A objectives, this study shifts focus from the average acquisition target to the entire distribution of M&A targets. We contribute to the literature by analyzing different takeover motivations in a multivariate setting. The high skewness of the deal value distribution strongly points at heterogeneity of acquisition targets and underlying motivations, so that the average effect might not be very meaningful. We add to the previous literature by leaving behind the concept of the average acquisition target and focus on the whole distribution of the deal value in order to investigate the importance of different merger objectives. Our results uncover different motivations underlying the acquisitions of SMEs and large firms. Using quantile regression (QR), we show how the importance of different firm characteristics changes along the deal value distribution. As opposed to standard regression techniques that focus on the average acquisition and average motivation QR provides a more complete picture of takeover motivations as it provides insights at different parts of the deal value distribution. Our empirical analysis is based on a comprehensive dataset of 652 European M&As in the period from 1997 to 2003. The results indicate that the importance of technological assets is indeed higher for smaller target firms while non-technological assets explain the largest share of the acquisition price of larger targets. The findings support the view of smaller acquisitions to complement the acquirer’s technology portfolio. Further, acquisitions of large firms tend to be used to gain access to international markets rather than small firm acquisitions. The remainder of the chapter is organized as follows. The next sections provide a short review on takeover motivations before we will outline our theoretical considerations and establish a set of hypotheses. Section 4 introduces the data set we use and presents descriptive statistics. The empirical test of our hypotheses is provided subsequently. Section 6 discusses our results and provides policy and managerial implications. The last section concludes with a critical evaluation of the study and points out potential areas for further research.

Market and Technology Access through Firm Acquisitions

293

A BRIEF REVIEW OF TAKEOVER MOTIVATIONS Previous literature has suggested that gaining access to technology and to geographical markets may be characterized as the two major motives for engaging in an M&A transaction. Technology-related M&As have received quite some attention by economists and management researchers in the recent past. Industrial organization economics put emphasis on market power and efficiency gains as major drivers of M&A activity (Caves, 1989; Scherer & Ross, 1990; Ro¨ller et al., 2001). These merger objectives can be taken forward to technology markets (Cassiman, Colombo, Garrone, & Veugelers, 2005). On the one hand, firm acquisitions are carried out to realize economies of scale and economies of scope in research and development (R&D) (Bertrand & Zungia, 2006; Cassiman et al., 2005). On the other hand, horizontal acquisitions may reduce competition and increase market power in technology markets (Chakrabarti, Hauschildt, & Su¨verkru¨p, 1994; Mukherjee, Kiymaz, & Baker, 2004). Complementary to this perspective, strategy researchers have argued that M&A transactions can be used to reconfigure the acquirer’s or target’s business, in order to respond to changes in the competitive environment or to enhance and improve existing operations (e.g., Bowman & Singh, 1993; Capron, Dussauge, & Mitchell, 1998; Capron & Hulland, 1999). Reconfiguring the business goes along with a redeployment of resources which, in case of R&D, may involve intellectual property rights, personnel, laboratories and technical instruments being physically transferred to new locations or used in different R&D projects. Such resource-based motivations for acquisitions have gained a lot of attention in the strategy literature. Moreover, the combination of two product or technology portfolios provides an opportunity to exploit complementarities (Ahuja & Katila, 2001; Colombo, Grilli, & Piva, 2006) that result from a skilled unbundling or bundling of resources with the objective to enhance (technological) core competencies of the merged entity (Cassiman et al., 2005; Sorescu, Chandy, & Prahbu, 2007). Besides technology, non-technological motivations may play a vital role in the decision to acquire a potential target. Those motivations tend to receive most importance in the trade-off between an acquisition and a greenfield investment (Go¨rg, 2000). In contrast to greenfield investments, acquisitions provide the acquiring firm almost immediately with an existing product portfolio customized to the market as well as with a developed distribution and customer network (Caves, 1989; Scherer & Ross, 1990; Ro¨ller et al., 2001). Moreover, competition in this market can be reduced. These benefits tend to be higher the larger the target firm is.

294

CHRISTOPH GRIMPE AND KATRIN HUSSINGER

In addition, M&As can be used to access foreign, i.e. not domestic, markets. Particularly in Europe, the past decades were characterized by an increased integration of markets as a consequence of the deepening of the European Union and the set-up of the European Monetary Union. After the abolishment of significant barriers to the free flow of capital, labor and trade between countries, M&As were, besides foreign direct investment, one of the major means to gain access to foreign European markets (Kleinert & Klodt, 2000). Foreign acquisition targets are attractive due to their familiarity with local consumer tastes, rules and the culture of the labor market, effective ways of advertising, the distribution network, government regulations, and market interactions between suppliers, consumers and competitors (Go¨rg, 2000; Qiu & Zhou, 2006).

ANALYTICAL FRAMEWORK Carrying out an acquisition involves the payment of a certain acquisition price, either for shares at the stock market or in a private sale. Acquiring firms will presumably pay a higher price for a target the more pronounced their takeover motivations are. Financial market efficiency suggests that the market value of a firm reflects the available information that relates to its current and future profitability (Fama, 1970). Jensen and Ruback (1983) have argued that acquisitions typically involve a significant positive premium over the market value of the target firm, which suggests that some acquirers put a higher value on the acquisition target than the market does. Previous literature provides a number of explanations for the attractiveness of certain targets beyond its actual market value ranging from their product market position to their intellectual property rights to their engagement in foreign markets. This section gives an overview on how these major merger motivations – technology and markets – translate into the acquisition price.

Technological Relatedness, Content and the Value of Technology Previous literature has shown that technological relatedness is a major factor for the success of M&As along with product market relatedness (Ahuja & Katila, 2001; Cassiman et al., 2005). In case of related strengths and core competencies acquiring firms can maximize complementarity effects from bundling strategic resources into unique and valuable

Market and Technology Access through Firm Acquisitions

295

combinations (Barney, 1991; Conner, 1991; Peteraf, 1993). Through this process of resource redeployment (Capron et al., 1998; Capron & Hulland, 1999), a merged entity may create a new or improved set of capabilities providing the basis for superior firm performance and competitive advantage (Penrose, 1959; Eisenhardt & Martin, 2000; Priem & Butler, 2001; Sorescu et al., 2007). Hence, acquiring firms presumably screen technology markets carefully in search for acquisition targets that will most effectively complement their technology portfolio. An important precondition for identifying valuable resources of a target firm is the ability of an investor to judge and value the potential of externally available technologies, which is summarized in the literature as the absorptive capacity of a firm (Cohen & Levinthal, 1989, 1990). Absorptive capacity is the ability to identify valuable technological knowledge in the environment, assimilate and finally exploit it in combination with existing know-how for successful innovation. In general, absorptive capacities increase awareness for market and technology trends, which can be translated into pre-emptive actions (Bowman & Hurry, 1993; Grimpe & Hussinger, 2008). As a result, they enable firms to predict future developments more accurately (Cohen & Levinthal, 1994). Acquirers who wish to realize complementarities have typically developed expertise and absorptive capacities as a by-product of R&D activities in a particular technology field (Cohen & Levinthal, 1989, 1990). An alternative theoretical perspective on the ability of an acquiring firm to judge the value of externally available technologies has emerged from the literature on information asymmetries in investment decisions (e.g., Aboody & Lev, 2000; Cohen & Dean, 2005; Heeley, Matusik, & Jain, 2007). Generally speaking, acquirers face the challenge of evaluating the innovation activities of a potential target in the absence of detailed information on every single innovation project. Each innovation project is characterized by its own specific attributes which are generally kept secret by a firm to appropriate the returns from innovation. In addition to their in-house expertise, acquirers may use publicly available information sources like patent data to assess the value of a firm’s innovation activities (Heeley et al., 2007). In order for a patent to be granted, the technological content of the patent needs to be disclosed by the applicant to the patent office. As it is highly technical information, providing only those ‘‘skilled in the art’’ with relevant knowledge about the true content, there is again substantial information asymmetry between informed and uninformed acquirers. This difference becomes even more pronounced when technological complexity increases, as is typical for high-technology industries (Hagedoorn & Duysters, 2002).

296

CHRISTOPH GRIMPE AND KATRIN HUSSINGER

Drawing from the theory of absorptive capacity and information asymmetries, we hence hypothesize that acquirers with technological expertise acquired through R&D activities in a particular field can better judge the value of a target’s technological assets. Furthermore, the combination of the acquiring firm’s technologies with related knowledge stocks of the target may lead to complementarity effects (Cassiman & Veugelers, 2006). Lastly, firms active in the same technology field as the acquirer may own patents and other intellectual property rights that block ongoing research within the acquiring company (O’Donoghue et al., 1998; Lerner et al., 2003). An acquisition can solve legal disputes over intellectual property rights and unlock blocked lines of research (Graff et al., 2003). Through a similar vein, firms can also engage in M&As to pre-empt competition in technology markets (Grimpe & Hussinger, 2008). Taking all those arguments together, our first hypothesis reads Hypothesis 1a. The price paid for an acquisition target increases if the acquiring and target firm are active in related technology fields. Furthermore, it is sensible to distinguish between the size of the technology portfolio to be acquired and its quality. In the first place, a patent acts as a positive signal as it shows that the prospective target firm has proven its technological expertise and capabilities and that it has a well-functioning laboratory and inventor team (Ndofor & Levitas, 2004; Levitas & McFadyen, 2006; Heeley et al., 2007). This holds also for acquiring firms that lack technological background or that engage in an acquisition in order to enter a new technology line. Moreover, patents have a value because they can be sold individually after the acquisition. Hence, our second hypothesis reads Hypothesis 2a. The price paid for an acquisition target increases with the target’s patent stock. Given the discussion on absorptive capacity, we argue that acquiring firms will also be able to identify valuable technological resources, i.e. highquality patents. Our third hypothesis hence reads Hypothesis 3a. The price paid for an acquisition target increases with the value of the target’s patents. Going beyond an average effect of technological assets, we hypothesize that information asymmetries are more pronounced for small firms (Shen & Reuer, 2005; Capron & Shen, 2007). Those firms are often privately held and young so that it is more difficult to access company information and

Market and Technology Access through Firm Acquisitions

297

especially information about their technological capabilities for outsiders to the technology field. Furthermore, the acquisition of small firms counteracts their potential of those firms to become strong competitors in future technology markets and the price paid for the acquisition target can outweigh this threat of future competition. From a transaction cost point of perspective (Williamson, 1985), the costs of exploiting new knowledge internally through firm acquisitions, as opposed to an external exploitation through technology licensing contracts at arm’s-lengths, has to outweigh the costs of the increase in governance. For small firms the acquisition and post-merger integration costs can be below technology licensing costs. Besides monetary costs, these costs are incurred through problems due to multiple ownership of patents that protect one technology in the presence of a fragmented technology space with mutual blocking patents (Heller & Eisenberg, 1998; Shapiro, 2001), value allocation and hold-up problems between the contracting parties, monitoring problems and strategic problems from rent-dissipating as a side effect of licensing that can also lead to new competitors (Graff et al., 2003; Ziedonis, 2004). Hence, the fact whether a potential acquisition target is operating in a related technology field will be more important for smaller targets as acquirers will be better able to evaluate the technologies employed. This will presumably result in a higher premium paid for those smaller targets relative to the premium paid to related but larger targets. Hypothesis 1b. The price paid for an acquisition target in related technology fields is relatively higher for M&As with a lower deal value than for those with a higher deal value. Given the discussion on transaction costs, we argue that technological assets are relatively more important for the acquisition of small firms because large firms are more likely to rely on engaging in technology licensing than in takeovers in the first place. This should be reflected in the size of the patent portfolio and its value. Hence, we hypothesize Hypothesis 2b. The price paid for an acquisition target’s patent stock is relatively higher for M&As with a lower deal value than for those with a higher deal value. Carrying forward the argument to the value of the acquired technology our last technology-related hypothesis reads Hypothesis 3b. The price paid for an acquisition target’s value of patents is relatively higher for M&As with a lower deal value than for those with a higher deal value.

298

CHRISTOPH GRIMPE AND KATRIN HUSSINGER

In the next section, we turn to the hypotheses on the importance of accessing foreign markets through firm acquisitions.

Entering Markets Through Firm Acquisition Under the increasing pressure of globalization the access to foreign markets is an important factor of today’s management strategy. Besides foreign direct investment, firm acquisitions across borders are an important means to access foreign markets (Go¨rg, 2000; Kleinert & Klodt, 2000), to acquire an existing distribution network and to benefit from the knowledge of local partners about different cultures and national market conditions. Furthermore, technology sourcing can play an important role for acquisitions across borders as the best possible match is not necessarily located in the same country (Frey & Hussinger, 2006; Sofka, 2007). The potential to transfer valuable intangible assets such as technological know-how between the merged firms is an important reasoning for crossmerger activities from a transaction point of perspective (Seth, Song, & Pettit, 2002). M&As across borders are however associated with higher information asymmetries than domestic M&As (Gioia & Thomsen, 2004) and hence with a higher risk of failure (Harris & Ravenscraft, 1991; Swenson, 1993). Transaction cost theory further suggests that transacting across country borders entails additional costs due to a different corporate culture, language barrier, and different national rules in labor and financial markets (Di Giovanni, 2005). To compensate for this additional risk cross-border mergers should generate higher expected benefits than a comparable domestic acquisition (Bertrand & Zungia, 2006). For this reason, we argue that expected gains from a cross-border merger are higher than from a comparable domestic M&A. This should be reflected in the deal value. Hence our fourth hypothesis reads Hypothesis 4a. The price paid for a cross-border acquisition target is higher than the price for a domestic target. Information asymmetries are supposed to be smaller for large established firms as financial and commercial data will be more readily available. Furthermore, the expected gains from M&As with large players are much higher. Hence we hypothesize

Market and Technology Access through Firm Acquisitions

299

Hypothesis 4b. The price paid for a cross-border acquisition target is relatively higher for M&As with a higher deal value than for those with a lower deal value. The next section explains the empirical model we use to test our hypothesis and introduces our data set.

METHODS Empirical Model The empirical model explains the deal value of the acquisition by the target firm’s assets and characteristics. As outlined above, our main focus is on the contribution of different merger motives found in the previous literature as access to technological assets and to foreign markets. To get insights into the effects of those different factors on the entire deal value distribution, we have to go beyond standard regression methods, which can only provide information on the average firm, and employ QR (Koenker & Bassett, 1978; Koenker & Hallock, 2000). In general, QR has several advantages. It accounts for the skewness of the deal value distribution, does not rely on a normality assumption of the error term of the regression model, and is robust to outliers. In contrast, ordinary least squares (OLS) regression is based on the mean of the conditional distribution of the deal value, which implicitly assumes that possible differences in terms of the impact of the exogenous variables along the conditional distribution are not existing or are unimportant. Focusing on the average firm only could hence hold back interesting information, especially in case of a skewed left-hand side variable as is the deal value. Acquisition targets with an extremely low or extremely high takeover price can deliver interesting information on different motivations underlying those M&As as opposed to the average transaction. In contrast to OLS, QR hence provides snapshots of different points of the deal value distribution. QR has the further advantage over OLS regression to be robust to outliers and heavy-tailed distributions (Buchinsky, 1994). In our setting targeting at a decomposition of the deal value, we define the acquired company in a hedonic way as a bundle of its characteristics and assets X (Hall, 1988; Gompers & Lerner, 2000). The deal value of the target V is a function of those characteristics X. In the presence of efficient markets and full information V(X) would equal the price at which the target firm’s assets are traded. Our empirical model then shows how the deal value is

300

CHRISTOPH GRIMPE AND KATRIN HUSSINGER

decomposed with respect to the target firm’s characteristics and assets V i ðX i Þ ¼ X i by þ uy

with

Quantiley ðV i jX i Þ ¼ X i by ,

(1)

where u is the error term of the empirical model, X the vector of exogenous variables and by the vector of coefficients. Quantiley (V|X) denotes the yth conditional quantile of V given X. The yth regression quantile, 0oyo1, is defined as a solution to the problem ( ) X X yjV i  X i bj þ ð1  yÞjV i  X i bj , (2) min b2